Tag Archives: machine learning

Doubling down on the edge with Coral’s new accelerator

Posted by The Coral Team

Coral image

Moving into the fall, the Coral platform continues to grow with the release of the M.2 Accelerator with Dual Edge TPU. Its first application is in Google’s Series One room kits where it helps to remove interruptions and makes the audio clearer for better video meetings. To help even more folks build products with Coral intelligence, we’re dropping the prices on several of our products. And for those folks that are looking to level up their at home video production, we’re sharing a demo of a pose based AI director to make multi-camera video easier to make.

Coral M.2 Accelerator with Dual Edge TPU

The newest addition to our product family brings two Edge TPU co-processors to systems in an M.2 E-key form factor. While the design requires a dual bus PCIe M.2 slot, it brings enhanced ML performance (8 TOPS) to tasks such as running two models in parallel or pipelining one large model across both Edge TPUs.

The ability to scale across multiple edge accelerators isn’t limited to only two Edge TPUs. As edge computing expands to local data centers, cell towers, and gateways, multi-Edge TPU configurations will be required to help process increasingly sophisticated ML models. Coral allows the use of a single toolchain to create models for one or more Edge TPUs that can address many different future configurations.

A great example of how the Coral M.2 Accelerator with Dual Edge TPU is being used is in the Series One meeting room kits for Google Meet.

The new Series One room kits for Google Meet run smarter with Coral intelligence

Coral image

Google’s new Series One room kits use our Coral M.2 Accelerator with Dual Edge TPU to bring enhanced audio clarity to video meetings. TrueVoice®, a multi-channel noise cancellation technology, minimizes distractions to ensure every voice is heard with up to 44 channels of echo and noise cancellation, making distracting sounds like snacking or typing on a keyboard a concern of the past.

Enabling the clearest possible communication in challenging environments was the target for the Google Meet hardware team. The consideration of what makes a challenging environment was not limited to unusually noisy environments, such as lunchrooms doubling as conference rooms. Any conference room can present challenging acoustics that make it difficult for all participants to be heard.

The secret to clarity without expensive and cumbersome equipment is to use virtual audio channels and AI driven sound isolation. Read more about how Coral was used to enhance and future-proof the innovative design.

Expanding the AI edge

Earlier this year, we reduced the prices of our prototyping devices and sensors. We are excited to share further price drops on more of our products. Our System-on-Module is now available for $99.99, and our Mini PCIe Accelerator, M.2 Accelerator A+E Key, and M.2 Accelerator B+M key are now available at $24.99. We hope this lower price will make our edge AI more accessible to more creative minds around the world. Later, this month our SoM offering will also expand to include 2 and 4GB RAM options.

Multi-cam with AI

Coral image

As we expand our platform and product family, we continue to keep new edge AI use cases in mind. We are continually inspired by our developer community’s experimentation and implementations. When recently faced with the challenges of multicam video production from home, Markku Lepistö, Solutions Architect at Google Cloud, created this real-time pose-based multicam tool he so aptly dubbed, AI Director.

We love seeing such unique implementations of on-device ML and invite you to share your own projects and feedback at [email protected].

For a list of worldwide distributors, system integrators and partners, visit the Coral partnerships page. Please visit Coral.ai to discover more about our edge ML platform.

Doubling down on the edge with Coral’s new accelerator

Posted by The Coral Team

Coral image

Moving into the fall, the Coral platform continues to grow with the release of the M.2 Accelerator with Dual Edge TPU. Its first application is in Google’s Series One room kits where it helps to remove interruptions and makes the audio clearer for better video meetings. To help even more folks build products with Coral intelligence, we’re dropping the prices on several of our products. And for those folks that are looking to level up their at home video production, we’re sharing a demo of a pose based AI director to make multi-camera video easier to make.

Coral M.2 Accelerator with Dual Edge TPU

The newest addition to our product family brings two Edge TPU co-processors to systems in an M.2 E-key form factor. While the design requires a dual bus PCIe M.2 slot, it brings enhanced ML performance (8 TOPS) to tasks such as running two models in parallel or pipelining one large model across both Edge TPUs.

The ability to scale across multiple edge accelerators isn’t limited to only two Edge TPUs. As edge computing expands to local data centers, cell towers, and gateways, multi-Edge TPU configurations will be required to help process increasingly sophisticated ML models. Coral allows the use of a single toolchain to create models for one or more Edge TPUs that can address many different future configurations.

A great example of how the Coral M.2 Accelerator with Dual Edge TPU is being used is in the Series One meeting room kits for Google Meet.

The new Series One room kits for Google Meet run smarter with Coral intelligence

Coral image

Google’s new Series One room kits use our Coral M.2 Accelerator with Dual Edge TPU to bring enhanced audio clarity to video meetings. TrueVoice®, a multi-channel noise cancellation technology, minimizes distractions to ensure every voice is heard with up to 44 channels of echo and noise cancellation, making distracting sounds like snacking or typing on a keyboard a concern of the past.

Enabling the clearest possible communication in challenging environments was the target for the Google Meet hardware team. The consideration of what makes a challenging environment was not limited to unusually noisy environments, such as lunchrooms doubling as conference rooms. Any conference room can present challenging acoustics that make it difficult for all participants to be heard.

The secret to clarity without expensive and cumbersome equipment is to use virtual audio channels and AI driven sound isolation. Read more about how Coral was used to enhance and future-proof the innovative design.

Expanding the AI edge

Earlier this year, we reduced the prices of our prototyping devices and sensors. We are excited to share further price drops on more of our products. Our System-on-Module is now available for $99.99, and our Mini PCIe Accelerator, M.2 Accelerator A+E Key, and M.2 Accelerator B+M key are now available at $24.99. We hope this lower price will make our edge AI more accessible to more creative minds around the world. Later, this month our SoM offering will also expand to include 2 and 4GB RAM options.

Multi-cam with AI

Coral image

As we expand our platform and product family, we continue to keep new edge AI use cases in mind. We are continually inspired by our developer community’s experimentation and implementations. When recently faced with the challenges of multicam video production from home, Markku Lepistö, Solutions Architect at Google Cloud, created this real-time pose-based multicam tool he so aptly dubbed, AI Director.

We love seeing such unique implementations of on-device ML and invite you to share your own projects and feedback at [email protected].

For a list of worldwide distributors, system integrators and partners, visit the Coral partnerships page. Please visit Coral.ai to discover more about our edge ML platform.

Announcing Google India AI/ML Research Awardees 2020

We recently concluded our annual edition of the Google India AI/ML Research Awards, a program focussed on supporting exceptional AI research in India. We also want to identify and strengthen long-term collaborative relationships with faculty working on problems that will impact how future generations use technology. 


This year we received over a hundred proposals across various fields of AI.  We also received several proposals working to advance the use of AI for Social Good. All proposals went through an extensive review process involving expert reviewers across Google who assessed the proposals on merit, innovation, connection to Google’s research efforts and alignment with our overall research philosophy and AI Principles


As a result, we are happy to announce our support for five faculty members who are working to  advance foundational and applied research to advance the state-of-the-art in AI across a wide range of research areas including Algorithms & Theory, Computer Vision, Natural Language Understanding and Privacy & Security. 

  • Arpita Patra, Associate Professor at the Indian Institute of Science is working on moulding the use of Secure Multiparty Computation (MPC) techniques to advance Machine Learning that preserves the privacy of the user, which can be tuned to real-world problems in the social good space, such as medical diagnosis systems, disparity against women, and fake news detection. 

  • Anirban Dasgupta, Associate Professor at the Indian Institute of Technology Gandhinagar is working on developing randomized approximation algorithms for numerical tensor algebra that strike a balance by being practically useful as well as by being equipped with theoretical guarantees. He also aims to develop such algorithms for applications such as streaming and large-scale social networks. 

  • Pawan Goyal, Associate Professor at the Indian Institute of Technology Kharagpur is developing ways to build conceptual understanding of natural language in AI Dialogue Systems. His work aims at developing dialog systems that can learn underlying concepts and perform commonsense reasoning to help AI systems in conversations.

  • Soma Biswas, Associate Professor at the Indian Institute of Science is working on making AI systems more robust by fundamentally advancing how deep learning algorithms recognize and directly provide information on what groups of data the system does not know much about . This work has widespread applications in image classification, detection, segmentation, etc.

  • Vasudeva Verma, Professor at International Institute of Information Technology Hyderabad is advancing his work on ‘Project ANGEL’, an initiative aimed at utilizing machine learning techniques for enhancing the well-being of teenagers, especially teenage girls. He intends to develop a cohesive technology stack (including prior work on Building on hate speech detection, sexism classification), through multi-disciplinary research for helping teenagers in an empathetic, proactive manner.


In the past we have supported various faculty members, including Sunita Sarawagi, working on continuously trainable learning systems with applications in grammar error correction and  translation. Our past awardee, Rijurekha Sen worked on developing low-cost, scalable measurement frameworks for real time monitoring of road traffic congestion and particulate matter in the air. 


We remain committed to investing in  the development of the Research ecosystem in India through various research grant-based and education programs, and continue to pursue cutting-edge research at Google Research India: an AI lab in Bangalore. More information about our program can be found here.


Posted by Ashwani Sharma, Head of Research Operations and University Relations, Google Research India,  and Divy Thakkar, Research and Education Program Manager


The Technology Behind our Recent Improvements in Flood Forecasting

Flooding is the most common natural disaster on the planet, affecting the lives of hundreds of millions of people around the globe and causing around $10 billion in damages each year. Building on our work in previous years, earlier this week we announced some of our recent efforts to improve flood forecasting in India and Bangladesh, expanding coverage to more than 250 million people, and providing unprecedented lead time, accuracy and clarity.

To enable these breakthroughs, we have devised a new approach for inundation modeling, called a morphological inundation model, which combines physics-based modeling with machine learning (ML) to create more accurate and scalable inundation models in real-world settings. Additionally, our new alert-targeting model allows identifying areas at risk of flooding at unprecedented scale using end-to-end machine learning models and data that is publicly available globally. In this post, we also describe developments for the next generation of flood forecasting systems, called HydroNets (presented at ICLR AI for Earth Sciences and EGU this year), which is a new architecture specially built for hydrologic modeling across multiple basins, while still optimizing for accuracy at each location.

Forecasting Water Levels
The first step in a flood forecasting system is to identify whether a river is expected to flood. Hydrologic models (or gauge-to-gauge models) have long been used by governments and disaster management agencies to improve the accuracy and extend the lead time of their forecasts. These models receive inputs like precipitation or upstream gauge measurements of water level (i.e., the absolute elevation of the water above sea level) and output a forecast for the water level (or discharge) in the river at some time in the future.

The hydrologic model component of the flood forecasting system described in this week’s Keyword post doubled the lead time of flood alerts for areas covering more than 75 million people. These models not only increase lead time, but also provide unprecedented accuracy, achieving an R2 score of more than 99% across all basins we cover, and predicting the water level within a 15 cm error bound more than 90% of the time. Once a river is predicted to reach flood level, the next step in generating actionable warnings is to convert the river level forecast into a prediction for how the floodplain will be affected.

Morphological Inundation Modeling
In prior work, we developed high quality elevation maps based on satellite imagery, and ran physics-based models to simulate water flow across these digital terrains, which allowed warnings with unprecedented resolution and accuracy in data-scarce regions. In collaboration with our satellite partners, Airbus, Maxar and Planet, we have now expanded the elevation maps to cover hundreds of millions of square kilometers. However, in order to scale up the coverage to such a large area while still retaining high accuracy, we had to re-invent how we develop inundation models.

Inundation modeling estimates what areas will be flooded and how deep the water will be. This visualization conceptually shows how inundation could be simulated, how risk levels could be defined (represented by red and white colors), and how the model could be used to identify areas that should be warned (green dots).

Inundation modeling at scale suffers from three significant challenges. Due to the large areas involved and the resolution required for such models, they necessarily have high computational complexity. In addition, most global elevation maps don’t include riverbed bathymetry, which is important for accurate modeling. Finally, the errors in existing data, which may include gauge measurement errors, missing features in the elevation maps, and the like, need to be understood and corrected. Correcting such problems may require collecting additional high-quality data or fixing erroneous data manually, neither of which scale well.

Our new approach to inundation modeling, which we call a morphological model, addresses these issues by using several innovative tricks. Instead of modeling the complex behaviors of water flow in real time, we compute modifications to the morphology of the elevation map that allow one to simulate the inundation using simple physical principles, such as those describing hydrostatic systems.

First, we train a pure-ML model (devoid of physics-based information) to estimate the one-dimensional river profile from gauge measurements. The model takes as input the water level at a specific point on the river (the stream gauge) and outputs the river profile, which is the water level at all points in the river. We assume that if the gauge increases, the water level increases monotonically, i.e., the water level at other points in the river increases as well. We also assume that the absolute elevation of the river profile decreases downstream (i.e., the river flows downhill).

We then use this learned model and some heuristics to edit the elevation map to approximately “cancel out” the pressure gradient that would exist if that region were flooded. This new synthetic elevation map provides the foundation on which we model the flood behavior using a simple flood-fill algorithm. Finally, we match the resulting flooded map to the satellite-based flood extent with the original stream gauge measurement.

This approach abandons some of the realistic constraints of classical physics-based models, but in data scarce regions where existing methods currently struggle, its flexibility allows the model to automatically learn the correct bathymetry and fix various errors to which physics-based models are sensitive. This morphological model improves accuracy by 3%, which can significantly improve forecasts for large areas, while also allowing for much more rapid model development by reducing the need for manual modeling and correction.

Alert targeting
Many people reside in areas that are not covered by the morphological inundation models, yet access to accurate predictions are still urgently needed. To reach this population and to increase the impact of our flood forecasting models, we designed an end-to-end ML-based approach, using almost exclusively data that is globally publicly available, such as stream gauge measurements, public satellite imagery, and low resolution elevation maps. We train the model to use the data it is receiving to directly infer the inundation map in real time.

A direct ML approach from real-time measurements to inundation.

This approach works well “out of the box” when the model only needs to forecast an event that is within the range of events previously observed. Extrapolating to more extreme conditions is much more challenging. Nevertheless, proper use of existing elevation maps and real-time measurements can enable alerts that are more accurate than presently available for those in areas not covered by the more detailed morphological inundation models. Because this model is highly scalable, we were able to launch it across India after only a few months of work, and we hope to roll it out to many more countries soon.

Improving Water Levels Forecasting
In an effort to continue improving flood forecasting, we have developed HydroNets — a specialized deep neural network architecture built specifically for water levels forecasting — which allows us utilize some exciting recent advances in ML-based hydrology in a real-world operational setting. Two prominent features distinguish it from standard hydrologic models. First, it is able to differentiate between model components that generalize well between sites, such as the modeling of rainfall-runoff processes, and those that are specific to a given site, like the rating curve, which converts a predicted discharge volume into an expected water level. This enables the model to generalize well to different sites, while still fine-tuning its performance to each location. Second, HydroNets takes into account the structure of the river network being modeled, by training a large architecture that is actually a web of smaller neural networks, each representing a different location along the river. This allows neural networks that are modeling upstream sites to pass information encoded in embeddings to models of downstream sites, so that every model can know everything it needs without a drastic increase in parameters.

The animation below illustrates the structure and flow of information in HydroNets. The output from the modeling of upstream sub-basins is combined into a single representation of a given basin state. It is then processed by the shared model component, which is informed by all basins in the network, and passed on to the label prediction model, which calculates the water level (and the loss function). The output from this iteration of the network is then passed on to inform downstream models, and so on.

An illustration of the HydroNets architecture.

We’re incredibly excited about this progress, and are working hard on improving our systems further.

Acknowledgements
This work is a collaboration between many research teams at Google, and is part of our AI for Social Good efforts. We'd also like to thank our Geo and Policy teams, as well as Google.org.

Source: Google AI Blog


KeyPose: Estimating the 3D Pose of Transparent Objects from Stereo

Estimating the position and orientation of 3D objects is one of the core problems in computer vision applications that involve object-level perception, such as augmented reality and robotic manipulation. In these applications, it is important to know the 3D position of objects in the world, either to directly affect them, or to place simulated objects correctly around them. While there has been much research on this topic using machine learning (ML) techniques, especially Deep Nets, most have relied on the use of depth sensing devices, such as the Kinect, which give direct measurements of the distance to an object. For objects that are shiny or transparent, direct depth sensing does not work well. For example, the figure below includes a number of objects (left), two of which are transparent stars. A depth device does not find good depth values for the stars, and gives a very poor reconstruction of the actual 3D points (right).

Left: RGB image of transparent objects.  Right: A four-panel image showing the reconstructed depth for the scene on the left.The top row includes depth images and the bottom row presents the 3D point cloud. The left panels were reconstructed using a depth camera and the right panels are output from the ClearGrasp model.  Note that although ClearGrasp inpaints the depth of the stars, it mistakes the actual depth of the rightmost one.

One solution to this problem, such as that proposed by ClearGrasp, is to use a deep neural network to inpaint the corrupted depth map of the transparent objects. Given a single RGB-D image of transparent objects, ClearGrasp uses deep convolutional networks to infer surface normals, masks of transparent surfaces, and occlusion boundaries, which it uses to refine the initial depth estimates for all transparent surfaces in the scene (far right in the figure above). This approach is very promising, and allows scenes with transparent objects to be processed by pose-estimation methods that rely on depth.  But inpainting can be tricky, especially when trained completely with synthetic images, and can still result in errors in depth.

In “KeyPose: Multi-View 3D Labeling and Keypoint Estimation for Transparent Objects”, presented at CVPR 2020 in collaboration with the Stanford AI Lab, we describe an ML system that estimates the depth of transparent objects by directly predicting 3D keypoints. To train the system we gather a large real-world dataset of images of transparent objects in a semi-automated way, and efficiently label their pose using 3D keypoints selected by hand. We then train deep models (called KeyPose) to estimate the 3D keypoints end-to-end from monocular or stereo images, without explicitly computing depth. The models work on objects both seen and unseen during training, for both individual objects and categories of objects. While KeyPose can work with monocular images, the extra information available from stereo images allows it to improve its results by a factor of two over monocular image input, with typical errors from 5 mm to 10 mm, depending on the objects. It substantially improves over state-of-the-art in pose estimation for these objects, even when competing methods are provided with ground truth depth. We are releasing the dataset of keypoint-labeled transparent objects for use by the research community.

Real-World Transparent Object Dataset with 3D Keypoint Labels
To facilitate gathering large quantities of real-world images, we set up a robotic data-gathering system in which a robot arm moves through a trajectory while taking video with two devices, a stereo camera and the Kinect Azure depth camera.

Automated image sequence capture using a robot arm with a stereo camera and an Azure Kinect device.

The AprilTags on the target enable accurate tracing of the pose of the cameras. By hand-labelling only a few images in each video with 2D keypoints, we can extract 3D keypoints for all frames of the video using multi-view geometry, thus increasing the labelling efficiency by a factor of 100.

We captured imagery for 15 different transparent objects in five categories, using 10 different background textures and four different poses for each object, yielding a total of 600 video sequences comprising 48k stereo and depth images. We also captured the same images with an opaque version of the object, to provide accurate ground truth depth images. All the images are labelled with 3D keypoints. We are releasing this dataset of real-world images publicly, complementing the synthetic ClearGrasp dataset with which it shares similar objects.

KeyPose Algorithm Using Early Fusion Stereo
The idea of using stereo images directly for keypoint estimation was developed independently for this project; it has also appeared recently in the context of hand-tracking. The diagram below shows the basic idea: the two images from a stereo camera are cropped around the object and fed to the KeyPose network, which predicts a sparse set of 3D keypoints that represent the 3D pose of the object. The network is trained using supervision from the labelled 3D keypoints.

One of the key aspects of stereo KeyPose is the use of early fusion to intermix the stereo images, and allow the network to implicitly compute disparity, in contrast to late fusion, in which keypoints are predicted for each image separately, and then combined. As shown in the diagram below, the output of KeyPose is a 2D keypoint heatmap in the image plane along with a disparity (i.e., inverse depth) heatmap for each keypoint. The combination of these two heatmaps yields the 3D coordinate of the keypoint, for each keypoint.

Keypose system diagram. Stereo images are passed to a CNN model to produce a probability heatmap for each keypoint.  This heatmap yields 2D image coordinates U,V for the keypoint.  The CNN model also produces a disparity (inverse depth) heatmap for each keypoint, which when combined with the U,V coordinates, gives a 3D position (X,Y,Z).

When compared to late fusion or to monocular input, early fusion stereo typically is twice as accurate.

Results
The images below show qualitative results of KeyPose on individual objects. On the left is one of the original stereo images; in the middle are the predicted 3D keypoints projected onto the image. On the right, we visualize points from a 3D model of the bottle, placed at the pose determined by the predicted 3D keypoints. The network is efficient and accurate, predicting keypoints with an MAE of 5.2 mm for the bottle and 10.1 mm for the mug using just 5 ms on a standard GPU.

The following table shows results for KeyPose on category-level estimation. The test set used a background texture not seen by the training set. Note that the MAE varies from 5.8 mm to 9.9 mm, showing the accuracy of the method.

Quantitative comparison of KeyPose with the state-of-the-art DenseFusion system, on category-level data. We provide DenseFusion with two versions of depth, one from the transparent objects, and one from opaque objects. <2cm is the percent of estimates with errors less than 2 cm. MAE is the mean absolute error of the keypoints, in mm.

For a complete accounting of quantitative results, as well as, ablation studies, please see the paper and supplementary materials and the KeyPose website.

Conclusion
This work shows that it is possible to accurately estimate the 3D pose of transparent objects from RGB images without reliance on depth images. It validates the use of stereo images as input to an early fusion deep net, where the network is trained to extract sparse 3D keypoints directly from the stereo pair. We hope the availability of an extensive, labelled dataset of transparent objects will help to advance the field. Finally, while we used semi-automatic methods to efficiently label the dataset, we hope to employ self-supervision methods in future work to do away with manual labelling.

Acknowledgements
I want to thank my co-authors, Xingyu Liu of Stanford University, and Rico Jonschkowski and Anelia Angelova; as well the many who helped us through discussions during the project and paper writing, including Andy Zheng, Shuran Song, Vincent Vanhoucke, Pete Florence, and Jonathan Tompson.

Source: Google AI Blog


ML Kit Pose Detection Makes Staying Active at Home Easier

Posted by Kenny Sulaimon, Product Manager, ML Kit; Chengji Yan and Areeba Abid, Software Engineers, ML Kit

ML Kit logo

Two months ago we introduced the standalone version of the ML Kit SDK, making it even easier to integrate on-device machine learning into mobile apps. Since then we’ve launched the Digital Ink Recognition API, and also introduced the ML Kit early access program. Our first two early access APIs were Pose Detection and Entity Extraction. We’ve received an overwhelming amount of interest in these new APIs and today, we are thrilled to officially add Pose Detection to the ML Kit lineup.

ML Kit Overview

A New ML Kit API, Pose Detection


Examples of ML Kit Pose Detection

ML Kit Pose Detection is an on-device, cross platform (Android and iOS), lightweight solution that tracks a subject's physical actions in real time. With this technology, building a one-of-a-kind experience for your users is easier than ever.

The API produces a full body 33 point skeletal match that includes facial landmarks (ears, eyes, mouth, and nose), along with hands and feet tracking. The API was also trained on a variety of complex athletic poses, such as Yoga positions.

Skeleton image detailing all 33 landmark points

Skeleton image detailing all 33 landmark points

Under The Hood

Diagram of the ML Kit Pose Detection Pipeline

The power of the ML Kit Pose Detection API is in its ease of use. The API builds on the cutting edge BlazePose pipeline and allows developers to build great experiences on Android and iOS, with little effort. We offer a full body model, support for both video and static image use cases, and have added multiple pre and post processing improvements to help developers get started with only a few lines of code.

The ML Kit Pose Detection API utilizes a two step process for detecting poses. First, the API combines an ultra-fast face detector with a prominent person detection algorithm, in order to detect when a person has entered the scene. The API is capable of detecting a single (highest confidence) person in the scene and requires the face of the user to be present in order to ensure optimal results.

Next, the API applies a full body, 33 landmark point skeleton to the detected person. These points are rendered in 2D space and do not account for depth. The API also contains a streaming mode option for further performance and latency optimization. When enabled, instead of running person detection on every frame, the API only runs this detector when the previous frame no longer detects a pose.

The ML Kit Pose Detection API also features two operating modes, “Fast” and “Accurate”. With the “Fast” mode enabled, you can expect a frame rate of around 30+ FPS on a modern Android device, such as a Pixel 4 and 45+ FPS on a modern iOS device, such as an iPhone X. With the “Accurate” mode enabled, you can expect more stable x,y coordinates on both types of devices, but a slower frame rate overall.

Lastly, we’ve also added a per point “InFrameLikelihood” score to help app developers ensure their users are in the right position and filter out extraneous points. This score is calculated during the landmark detection phase and a low likelihood score suggests that a landmark is outside the image frame.

Real World Applications


Examples of a pushup and squat counter using ML Kit Pose Detection

Keeping up with regular physical activity is one of the hardest things to do while at home. We often rely on gym buddies or physical trainers to help us with our workouts, but this has become increasingly difficult. Apps and technology can often help with this, but with existing solutions, many app developers are still struggling to understand and provide feedback on a user’s movement in real time. ML Kit Pose Detection aims to make this problem a whole lot easier.

The most common applications for Pose detection are fitness and yoga trackers. It’s possible to use our API to track pushups, squats and a variety of other physical activities in real time. These complex use cases can be achieved by using the output of the API, either with angle heuristics, tracking the distance between joints, or with your own proprietary classifier model.

To get you jump started with classifying poses, we are sharing additional tips on how to use angle heuristics to classify popular yoga poses. Check it out here.

Learning to Dance Without Leaving Home

Learning a new skill is always tough, but learning to dance without the aid of a real time instructor is even tougher. One of our early access partners, Groovetime, has set out to solve this problem.

With the power of ML Kit Pose Detection, Groovetime allows users to learn their favorite dance moves from popular short-form dance videos, while giving users automated real time feedback on their technique. You can join their early access beta here.

Groovetime App using ML Kit Pose Detection

Staying Active Wherever You Are

Our Pose Detection API is also helping adidas Training, another one of our early access partners, build a virtual workout experience that will help you stay active no matter where you are. This one-of-a-kind innovation will help analyze and give feedback on the user’s movements, using nothing more than just your phone. Integration into the adidas Training app is still in the early phases of the development cycle, but stay tuned for more updates in the future.

How to get started?

If you would like to start using the Pose Detection API in your mobile app, head over to the developer documentation or check out the sample apps for Android and iOS to see the API in action. For questions or feedback, please reach out to us through one of our community channels.

Understanding View Selection for Contrastive Learning

Most people take for granted the ability to view an object from several different angles, but still recognize that it's the same object— a dog viewed from the front is still a dog when viewed from the side. While people do this naturally, computer scientists need to explicitly enable machines to learn representations that are view-invariant, with the goal of seeking robust data representations that retain information that is useful to downstream tasks.

Of course, in order to learn these representations, manually annotated training data can be used. However, as in many cases such annotations aren’t available, which gives rise to a series of self- and crossmodal supervised approaches that do not require manually annotated training data. Currently, a popular paradigm for training with such data is contrastive multiview learning, where two views of the same scene (for example, different image channels, augmentations of the same image, and video and text pairs) will tend to converge in representation space while two views of different scenes diverge. Despite their success, one important question remains: “If one doesn’t have annotated labels readily available, how does one select the views to which the representations should be invariant?” In other words, how does one identify an object using information that resides in the pixels of the image itself, while still remaining accurate when that image is viewed from disparate viewpoints?

In “What makes for good views for contrastive learning”, we use theoretical and empirical analysis to better understand the importance of view selection, and argue that one should reduce the mutual information between views while keeping task-relevant information intact. To verify this hypothesis, we devise unsupervised and semi-supervised frameworks that learn effective views by aiming to reduce their mutual information. We also consider data augmentation as a way to reduce mutual information, and show that increasing data augmentation indeed leads to decreasing mutual information while improving downstream classification accuracy. To encourage further research in this space, we have open-sourced the code and pre-trained models.

The InfoMin Hypothesis
The goal of contrastive multiview learning is to learn a parametric encoder, whose output representations can be used to discriminate between pairs of views with the same identities, and pairs with different identities. The amount and type of information shared between the views determines how well the resulting model performs on downstream tasks. We hypothesize that the views that yield the best results should discard as much information in the input as possible except for the task relevant information (e.g., object labels), which we call the InfoMin principle.

Consider the example below in which two patches of the same image represent the different “views”. The training objective is to identify that the two views belong to the same image. It is undesirable to have views that share too much information, for example, where low-level color and texture cues can be exploited as “shortcuts” (left), or to have views that share too little information to identify that they belong to the same image (right). Rather, views at the “sweet spot” share the information related to downstream tasks, such as patches corresponding to different parts of the panda for an object classification task (center).

An illustration of three regimes of information captured during contrastive multiview learning. Views should not share too much information (left) or too little information (right), but should find an optimal mix (the “sweet spot”, middle) that maximizes the downstream performance.

A Unified View on Contrastive Learning
We design several sets of experiments to verify the InfoMin hypothesis, motivated by the fact that there are simple ways to control the mutual information shared between views without any supervision. For example, we can sample different patches from the same images, and reduce their mutual information simply by increasing the distance between the patches. Here, we estimate the mutual information using InfoNCE (INCE), which is a quantitative measure of the mutual information lower bound.. Indeed, we observe a reverse U-shape curve: as mutual information is reduced, the downstream task accuracy first increases and then begins to decrease.

Downstream classification accuracy on STL-10 (left) and CIFAR-10 (right) by applying linear classifiers on representations learned with contrastive learning. Same as the previous illustration, the views are sampled as different patches from the same images. Increasing the Euclidean distance between patches leads to decreasing mutual information. A reverse U-shape curve between classification accuracy and INCE (patch distance) is observed.

Furthermore, we demonstrate that several state-of-the-art contrastive learning methods (InstDis, MoCo, CMC, PIRL, SimCLR and CPC) can be unified through the perspective of view selection: despite the differences in architecture, objective and engineering details, all recent contrastive learning methods create two views that implicitly follow the InfoMin hypothesis, where the information shared between views are controlled by the strength of data augmentation. Motivated by this, we propose a new set of data augmentations, which outperforms the prior state of the art, SimCLR, by nearly 4% on the ImageNet linear readout benchmark. We also found that transferring our unsupervised pre-trained models to object detection and instance segmentation consistently outperforms ImageNet pre-training.

Learning to Generate Views
In our work, we design unsupervised and semi-supervised methods that synthesize novel views following the InfoMin hypothesis. We learn flow-based models that transfer natural color spaces into novel color spaces, from which we split the channels to get views. For the unsupervised setup, the view generators are optimized to minimize the InfoNCE bound between views. As shown in the results below, we observe a similar reverse U-shape trend while minimizing the InfoNCE bound.

View generators learned by unsupervised (left) and semi-supervised (right) objectives.

To reach the sweet spot without overly minimizing mutual information, we can use the semi-supervised setup and guide the view generator to retain label information. As expected, all learned views are now centered around the sweet spot, no matter what the input color space is.

Code and Pretrained Models
To accelerate research in self-supervised contastive learning, we are excited to share the code and pretrained models of InfoMin with the academic community. They can be found here.

Acknowledgements
The core team includes Yonglong Tian, Chen Sun, Ben Poole, Dilip Krishnan, Cordelia Schmid and Phillip Isola. We would like to thank Kevin Murphy for insightful discussion; Lucas Beyer for feedback on the manuscript; and the Google Cloud team for computation support.

Source: Google AI Blog


Introducing TensorFlow Recorder

When training computer vision machine learning models, data loading can often be a performance bottleneck, causing your GPU or TPU resources to be underutilized while waiting for data to be loaded into the model. Storing your dataset in the efficient TensorFlow Record (TFRecord) format is a great way to solve these problems, but creating TFRecords can unfortunately often require a great deal of complex code.

Last week we open sourced the TensorFlow Recorder project (also known as TFRecorder), which makes it possible for data scientists, data engineers, or AI/ML engineers to create image based TFRecords with just a few lines of code. Using TFRecords is incredibly important for creating efficient TensorFlow ML pipelines, but until now they haven’t been so easy to create. Before TFRecorder, in order to create TFRecords at scale you would have had to write a data pipeline that parsed your structured data, loaded images from storage, and serialized the results into the TFRecord format. TFRecorder allows you to write TFRecords directly from a Pandas dataframe or CSV without writing any complicated code.

You can see an example of TFRecoder below, but first let’s talk about some of the specific advantages of TFRecords.

How TFRecords Can Help

Using the TFRecord file format allows you to store your data in sets of files, each containing a sequence of protocol buffers serialized as a binary record that can be read very efficiently, which will help reduce the data loading bottleneck mentioned above.

Data loading performance can be further improved by implementing prefetching and parallel interleave along with using the TFRecord format. Prefetching reduces the time of each model training step(s) by fetching the data for the next training step while your model is executing training on the current step. Parallel interleave allows you to read from multiple TFRecords shards (pieces of a TFRecord file) and apply preprocessing of those interleaved data streams. This reduces the latency required to read a training batch and is especially helpful when reading data from the network.

Using TensorFlow Recorder

Creating a TFRecord using TFRecorder requires only a few lines of code. Here’s how it works.
import pandas as pd
import tfrecorder
df = pd.read_csv(...)
df.tensorflow.to_tfrecord(output_dir="gs://my/bucket")

TFRecorder currently expects data to be in the same format as Google AutoML Vision.

This format looks like a pandas dataframe or CSV formatted as:
splitimage_urilabel
TRAIN
gs://my/bucket/image1.jpgcat

Where:
  • split can take on the values TRAIN, VALIDATION, and TEST
  • image_uri specifies a local or google cloud storage location for the image file.
  • label can be either a text-based label that will be integerized or an integer
In the future, we hope to extend TensorFlow Recorder to work with data in any format.

While this example would work well to convert a few thousand images into TFRecords, it probably wouldn’t scale well if you have millions of images. To scale up to huge datasets, TensorFlow Recorder provides connectivity with Google Cloud Dataflow, which is a serverless Apache Beam pipeline runner. Scaling up to DataFlow requires only a little bit more configuration.
df.tensorflow.to_tfrecord(
output_dir="gs://my/bucket",
runner="DataFlowRunner",
project="my-project",
region="us-central1)

What’s next?

We’d love for you to try out TensorFlow Recorder. You can get it from GitHub or simply pip install tfrecorder. Tensorflow Recorder is very new and we’d greatly appreciate your feedback, suggestions, and pull requests.

By Mike Bernico and Carlos Ezequiel, Google Cloud AI Engineers

Announcing ScaNN: Efficient Vector Similarity Search



Suppose one wants to search through a large dataset of literary works using queries that require an exact match of title, author, or other easily machine-indexable criteria. Such a task would be well suited for a relational database using a language such as SQL. However, if one wants to support more abstract queries, such as “Civil War poem,” it is no longer possible to rely on naive similarity metrics such as the number of words in common between two phrases. For example, the query “science fiction” is more related to “future” than it is to “earth science” despite the former having zero, and the latter having one, word in common with the query.

Machine learning (ML) has greatly improved computers’ abilities to understand language semantics and therefore answer these abstract queries. Modern ML models can transform inputs such as text and images into embeddings, high dimensional vectors trained such that more similar inputs cluster closer together. For a given query, we can therefore compute its embedding, and find the literary works whose embeddings are closest to the query’s. In this manner, ML has transformed an abstract and previously difficult-to-specify task into a rigorous mathematical one. However, a computational challenge remains: for a given query embedding, how does one quickly find the nearest dataset embeddings? The set of embeddings is often too large for exhaustive search and its high dimensionality makes pruning difficult.

In our ICML 2020 paper, “Accelerating Large-Scale Inference with Anisotropic Vector Quantization,” we address this problem by focusing on how to compress the dataset vectors to enable fast approximate distance computations, and propose a new compression technique that significantly boosts accuracy compared to prior works. This technique is utilized in our recently open-sourced vector similarity search library (ScaNN), and enables us to outperform other vector similarity search libraries by a factor of two, as measured on ann-benchmarks.com.

The Importance of Vector Similarity Search
Embedding-based search is a technique that is effective at answering queries that rely on semantic understanding rather than simple indexable properties. In this technique, machine learning models are trained to map the queries and database items to a common vector embedding space, such that the distance between embeddings carries semantic meaning, i.e., similar items are closer together.
The two-tower neural network model, illustrated above, is a specific type of embedding-based search where queries and database items are mapped to the embedding space by two respective neural networks. In this example the model responds to natural-language queries for a hypothetical literary database.
To answer a query with this approach, the system must first map the query to the embedding space. It then must find, among all database embeddings, the ones closest to the query; this is the nearest neighbor search problem. One of the most common ways to define the query-database embedding similarity is by their inner product; this type of nearest neighbor search is known as maximum inner-product search (MIPS).

Because the database size can easily be in the millions or even billions, MIPS is often the computational bottleneck to inference speed, and exhaustive search is impractical. This necessitates the use of approximate MIPS algorithms that exchange some accuracy for a significant speedup over brute-force search.

A New Quantization Approach for MIPS
Several state-of-the-art solutions for MIPS are based on compressing the database items so that an approximation of their inner product can be computed in a fraction of the time taken by brute-force. This compression is commonly done with learned quantization, where a codebook of vectors is trained from the database and is used to approximately represent the database elements.

Previous vector quantization schemes quantized database elements with the aim of minimizing the average distance between each vector x and its quantized form . While this is a useful metric, optimizing for this is not equivalent to optimizing nearest-neighbor search accuracy. The key idea behind our paper is that encodings with higher average distance may actually result in superior MIPS accuracy.

The intuition for our result is illustrated below. Suppose we have two database embeddings x1 and x2, and must quantize each to one of two centers: c1 or c2. Our goal is to quantize each xi to i such that the inner product <q, i> is as similar to the original inner product <q, xi> as possible. This can be visualized as making the magnitude of the projection of i onto q as similar as possible to the projection of xi onto q. In the traditional approach to quantization (left), we would pick the closest center for each xi, which leads to an incorrect relative ranking of the two points: <q, 1> is greater than <q, 2>, even though <q, x1> is less than <q, x2>! If we instead assign x1 to c1 and x2 to c2, we get the correct ranking. This is illustrated in the figure below.
The goal is to quantize each xi to i = c1 or i = c2. Traditional quantization (left) results in the incorrect ordering of x1 and x2 for this query. Even though our approach (right) chooses centers farther away from the data points, this in fact leads to lower inner product error and higher accuracy.
It turns out that direction matters as well as magnitude--even though c1 is farther from x1 than c2, c1 is offset from x1 in a direction almost entirely orthogonal to x1, while c2’s offset is parallel (for x2, the same situation applies but flipped). Error in the parallel direction is much more harmful in the MIPS problem because it disproportionately impacts high inner products, which by definition are the ones that MIPS is trying to estimate accurately.

Based on this intuition, we more heavily penalize quantization error that is parallel to the original vector. We refer to our novel quantization technique as anisotropic vector quantization due to the directional dependence of its loss function. The ability of this technique to trade increased quantization error of lower inner products in exchange for superior accuracy for high inner products is the key innovation and the source of its performance gains.
In the above diagrams, ellipses denote contours of equal loss. In anisotropic vector quantization, error parallel to the original data point x is penalized more.
Anisotropic Vector Quantization in ScaNN
Anisotropic vector quantization allows ScaNN to better estimate inner products that are likely to be in the top-k MIPS results and therefore achieve higher accuracy. On the glove-100-angular benchmark from ann-benchmarks.com, ScaNN outperformed eleven other carefully tuned vector similarity search libraries, handling roughly twice as many queries per second for a given accuracy as the next-fastest library.*
[email protected] is a commonly used metric for nearest neighbor search accuracy, which measures the proportion of the true nearest k neighbors that are present in an algorithm’s returned k neighbors. ScaNN (upper purple line) consistently achieves superior performance across various points of the speed-accuracy trade-off.
ScaNN is open-source software and you can try it yourself at GitHub. The library can be directly installed via Pip and has interfaces for both TensorFlow and Numpy inputs. Please see the GitHub repository for further instructions on installing and configuring ScaNN.

Conclusion
By modifying the vector quantization objective to align with the goals of MIPS, we achieve state-of-the-art performance on nearest neighbor search benchmarks, a key indicator of embedding-based search performance. Although anisotropic vector quantization is an important technique, we believe it is just one example of the performance gains achievable by optimizing algorithms for the end goal of improving search accuracy rather than an intermediate goal such as compression distortion.

Acknowledgements
This post reflects the work of the entire ScaNN team: David Simcha, Erik Lindgren, Felix Chern, Nathan Cordeiro, Ruiqi Guo, Sanjiv Kumar, and Zonglin Li. We’d also like to thank Dan Holtmann-Rice, Dave Dopson, and Felix Yu.



* ScaNN performs similarly well on the other datasets of ann-benchmarks.com, but the website currently shows outdated, lower numbers. See this pull request for more representative performance figures on other datasets.

Source: Google AI Blog


Summer updates from Coral

Posted by the Coral Team

Summer has arrived along with a number of Coral updates. We're happy to announce a new partnership with balena that helps customers build, manage, and deploy IoT applications at scale on Coral devices. In addition, we've released a series of updates to expand platform compatibility, make development easier, and improve the ML capabilities of our devices.

Open-source Edge TPU runtime now available on GitHub

First up, our Edge TPU runtime is now open-source and available on GitHub, including scripts and instructions for building the library for Linux and Windows. Customers running a platform that is not officially supported by Coral, including ARMv7 and RISC-V can now compile the Edge TPU runtime themselves and start experimenting. An open source runtime is easier to integrate into your customized build pipeline, enabling support for creating Yocto-based images as well as other distributions.

Windows drivers now available for the Mini PCIe and M.2 accelerators

Coral customers can now also use the Mini PCIe and M.2 accelerators on the Microsoft Windows platform. New Windows drivers for these products complement the previously released Windows drivers for the USB accelerator and make it possible to start prototyping with the Coral USB Accelerator on Windows and then to move into production with our Mini PCIe and M.2 products.

New fresh bits on the Coral ML software stack

We’ve also made a number of new updates to our ML tools:

  • The Edge TPU compiler is now version 14.1. It can be updated by running sudo apt-get update && sudo apt-get install edgetpu, or follow the instructions here
  • Our new Model Pipelining API allows you to divide your model across multiple Edge TPUs. The C++ version is currently in beta and the source is on GitHub
  • New embedding extractor models for EfficientNet, for use with on-device backpropagation. Embedding extractor models are compiled with the last fully-connected layer removed, allowing you to retrain for classification. Previously, only Inception and MobileNet were available and now retraining can also be done on EfficientNet
  • New Colab notebooks to retrain a classification model with TensorFlow 2.0 and build C++ examples

Balena partners with Coral to enable AI at the edge

We are excited to share that the Balena fleet management platform now supports Coral products!

Companies running a fleet of ML-enabled devices on the edge need to keep their systems up-to-date with the latest security patches in order to protect data, model IP and hardware from being compromised. Additionally, ML applications benefit from being consistently retrained to recognize new use cases with maximum accuracy. Coral + balena together, bring simplicity and ease to the provisioning, deployment, updating, and monitoring of your ML project at the edge, moving early prototyping seamlessly towards production environments with many thousands of devices.

Read more about all the benefits of Coral devices combined with balena container technology or get started deploying container images to your Coral fleet with this demo project.

New version of Mendel Linux

Mendel Linux (5.0 release Eagle) is now available for the Coral Dev Board and SoM and includes a more stable package repository that provides a smoother updating experience. It also brings compatibility improvements and a new version of the GPU driver.

New models

Last but not least, we’ve recently released BodyPix, a Google person-segmentation model that was previously only available for TensorFlow.JS, as a Coral model. This enables real-time privacy preserving understanding of where people (and body parts) are on a camera frame. We first demoed this at CES 2020 and it was one of our most popular demos. Using BodyPix we can remove people from the frame, display only their outline, and aggregate over time to see heat maps of population flow.

Here are two possible applications of BodyPix: Body-part segmentation and anonymous population flow. Both are running on the Dev Board.

We’re excited to add BodyPix to the portfolio of projects the community is using to extend our models far beyond our demos—including tackling today’s biggest challenges. For example, Neuralet has taken our MobileNet V2 SSD Detection model and used it to implement Smart Social Distancing. Using the bounding box of person detection, they can compute a region for safe distancing and let a user know if social distance isn’t being maintained. The best part is this is done without any sort of facial recognition or tracking, with Coral we can accomplish this in real-time in a privacy preserving manner.

We can’t wait to see more projects that the community can make with BodyPix. Beyond anonymous population flow there’s endless possibilities with background and body part manipulation. Let us know what you come up with at our community channels, including GitHub and StackOverflow.

________________________

We are excited to share all that Coral has to offer as we continue to evolve our platform. For a list of worldwide distributors, system integrators and partners, including balena, visit the Coral partnerships page. Please visit Coral.ai to discover more about our edge ML platform and share your feedback at [email protected].