In this paper we report the set-up and results of the Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) organized in conjunction with the MICCAI 2012 and 2013 conferences Twenty state-of-the-art tumor segmentation algorithms were applied to a set of 65 multi-contrast MR scans of low- and high-grade glioma patients—manually annotated by up to four raters—and to 65 comparable scans generated using tumor image simulation software Quantitative evaluations revealed considerable disagreement between the human raters in segmenting various tumor sub-regions (Dice scores in the range 74%–85%), illustrating the difficulty of this task We found that different algorithms worked best for different sub-regions (reaching performance comparable to human inter-rater variability), but that no single algorithm ranked in the top for all sub-regions simultaneously Fusing several good algorithms using a hierarchical majority vote yielded segmentations that consistently ranked above all individual algorithms, indicating remaining opportunities for further methodological improvements The BRATS image data and manual annotations continue to be publicly available through an online evaluation system as an ongoing benchmarking resource

/pdf/the-multimodal-brain-tumor-image-segmentation-benchmark-2gfraj98dk.pdf

The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS)

Dynamics of human body skeletons convey significant information for human action recognition. Conventional approaches for modeling skeletons usually rely on hand-crafted parts or traversal rules, thus resulting in limited expressive power and difficulties of generalization. In this work, we propose a novel model of dynamic skeletons called Spatial-Temporal Graph Convolutional Networks (ST-GCN), which moves beyond the limitations of previous methods by automatically learning both the spatial and temporal patterns from data. This formulation not only leads to greater expressive power but also stronger generalization capability. On two large datasets, Kinetics and NTU-RGBD, it achieves substantial improvements over mainstream methods.

/pdf/spatial-temporal-graph-convolutional-networks-for-skeleton-1mh8rphegh.pdf

Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition

Recent advances in 3D depth cameras such as Microsoft Kinect sensors (www.xbox.com/en-US/kinect) have created many opportunities for multimedia computing. The Kinect sensor lets the computer directly sense the third dimension (depth) of the players and the environment. It also understands when users talk, knows who they are when they walk up to it, and can interpret their movements and translate them into a format that developers can use to build new experiences. While the Kinect sensor incorporates several advanced sensing hardware, this article focuses on the vision aspect of the Kinect sensor and its impact beyond the gaming industry.

/pdf/microsoft-kinect-sensor-and-its-effect-m2gqa92s3h.pdf

Microsoft Kinect Sensor and Its Effect

We introduce a new dataset, Human3.6M, of 3.6 Million accurate 3D Human poses, acquired by recording the performance of 5 female and 6 male subjects, under 4 different viewpoints, for training realistic human sensing systems and for evaluating the next generation of human pose estimation models and algorithms. Besides increasing the size of the datasets in the current state-of-the-art by several orders of magnitude, we also aim to complement such datasets with a diverse set of motions and poses encountered as part of typical human activities (taking photos, talking on the phone, posing, greeting, eating, etc.), with additional synchronized image, human motion capture, and time of flight (depth) data, and with accurate 3D body scans of all the subject actors involved. We also provide controlled mixed reality evaluation scenarios where 3D human models are animated using motion capture and inserted using correct 3D geometry, in complex real environments, viewed with moving cameras, and under occlusion. Finally, we provide a set of large-scale statistical models and detailed evaluation baselines for the dataset illustrating its diversity and the scope for improvement by future work in the research community. Our experiments show that our best large-scale model can leverage our full training set to obtain a 20% improvement in performance compared to a training set of the scale of the largest existing public dataset for this problem. Yet the potential for improvement by leveraging higher capacity, more complex models with our large dataset, is substantially vaster and should stimulate future research. The dataset together with code for the associated large-scale learning models, features, visualization tools, as well as the evaluation server, is available online at http://vision.imar.ro/human3.6m .

/pdf/human3-6m-large-scale-datasets-and-predictive-methods-for-3d-3g2xlb8fic.pdf

Human3.6M: Large Scale Datasets and Predictive Methods for 3D Human Sensing in Natural Environments

Providing accurate and opportune information on people's activities and behaviors is one of the most important tasks in pervasive computing. Innumerable applications can be visualized, for instance, in medical, security, entertainment, and tactical scenarios. Despite human activity recognition (HAR) being an active field for more than a decade, there are still key aspects that, if addressed, would constitute a significant turn in the way people interact with mobile devices. This paper surveys the state of the art in HAR based on wearable sensors. A general architecture is first presented along with a description of the main components of any HAR system. We also propose a two-level taxonomy in accordance to the learning approach (either supervised or semi-supervised) and the response time (either offline or online). Then, the principal issues and challenges are discussed, as well as the main solutions to each one of them. Twenty eight systems are qualitatively evaluated in terms of recognition performance, energy consumption, obtrusiveness, and flexibility, among others. Finally, we present some open problems and ideas that, due to their high relevance, should be addressed in future research.

A Survey on Human Activity Recognition using Wearable Sensors

We propose a new method to quickly and accurately predict 3D positions of body joints from a single depth image, using no temporal information. We take an object recognition approach, designing an intermediate body parts representation that maps the difficult pose estimation problem into a simpler per-pixel classification problem. Our large and highly varied training dataset allows the classifier to estimate body parts invariant to pose, body shape, clothing, etc. Finally we generate confidence-scored 3D proposals of several body joints by reprojecting the classification result and finding local modes. The system runs at 200 frames per second on consumer hardware. Our evaluation shows high accuracy on both synthetic and real test sets, and investigates the effect of several training parameters. We achieve state of the art accuracy in our comparison with related work and demonstrate improved generalization over exact whole-skeleton nearest neighbor matching.

/pdf/real-time-human-pose-recognition-in-parts-from-single-depth-16ej02o3m1.pdf

Real-time human pose recognition in parts from single depth images

A system and method are disclosed for recognizing and tracking a user's skeletal joints with a NUI system and further, for recognizing and tracking only some skeletal joints, such as for example a user's upper body The system may include a limb identification engine which may use various methods to evaluate, identify and track positions of body parts of one or more users in a scene In examples, further processing efficiency may be achieved by segmenting the field of view in smaller zones, and focusing on one zone at a time Moreover, each zone may have its own set of predefined gestures which are recognized

Skeletal joint recognition and tracking system

Synthesized body images are generated for a machine learning algorithm of a body joint tracking system. Frames from motion capture sequences are retargeted to several different body types, to leverage the motion capture sequences. To avoid providing redundant or similar frames to the machine learning algorithm, and to provide a compact yet highly variegated set of images, dissimilar frames can be identified using a similarity metric. The similarity metric is used to locate frames which are sufficiently distinct, according to a threshold distance. For realism, noise is added to the depth images based on noise sources which a real world depth camera would often experience. Other random variations can be introduced as well. For example, a degree of randomness can be added to retargeting. For each frame, the depth image and a corresponding classification image, with labeled body parts, are provided. 3-D scene elements can also be provided.

Proxy training data for human body tracking

Technology is described for determining and using invariant features for computer vision. A local orientation may be determined for each depth pixel in a subset of the depth pixels in a depth map. The local orientation may an in-plane orientation, an out-out-plane orientation or both. A local coordinate system is determined for each of the depth pixels in the subset based on the local orientation of the corresponding depth pixel. A feature region is defined relative to the local coordinate system for each of the depth pixels in the subset. The feature region for each of the depth pixels in the subset is transformed from the local coordinate system to an image coordinate system of the depth map. The transformed feature regions are used to process the depth map.

Invariant features for computer vision

A computerized decision tree training system may include a distributed control processing unit configured to receive input of training data for training a decision tree. The system may further include a plurality of data batch processing units, each data batch processing unit being configured to evaluate each of a plurality of split functions of a decision tree for respective data batch of the training data, to thereby compute a partial histogram for each split function, for each datum in the data batch. The system may further include a plurality of node batch processing units configured to aggregate the associated partial histograms for each split function to form an aggregated histogram for each split function for each of a subset of frontier tree nodes and to determine a selected split function for each frontier tree node by computing the split function that produces highest information gain for the frontier tree node.

Richard E. Moore

Papers

Real-time human pose recognition in parts from single depth images

Skeletal joint recognition and tracking system

Proxy training data for human body tracking

Invariant features for computer vision

Distributed decision tree training