Over the years, a large number of methods have been proposed to analyze human pose and motion information from images, videos, and recently from depth data. Most methods, however, have been evaluated on datasets that were too specific to each application, limited to a particular modality, and more importantly, captured under unknown conditions. To address these issues, we introduce the Berkeley Multimodal Human Action Database (MHAD) consisting of temporally synchronized and geometrically calibrated data from an optical motion capture system, multi-baseline stereo cameras from multiple views, depth sensors, accelerometers and microphones. This controlled multimodal dataset provides researchers an inclusive testbed to develop and benchmark new algorithms across multiple modalities under known capture conditions in various research domains. To demonstrate possible use of MHAD for action recognition, we compare results using the popular Bag-of-Words algorithm adapted to each modality independently with the results of various combinations of modalities using the Multiple Kernel Learning. Our comparative results show that multimodal analysis of human motion yields better action recognition rates than unimodal analysis.

In this paper, we introduce Recipe1M+, a new large-scale, structured corpus of over one million cooking recipes and 13 million food images. As the largest publicly available collection of recipe data, Recipe1M+ affords the ability to train high-capacity models on aligned, multi-modal data. Using these data, we train a neural network to learn a joint embedding of recipes and images that yields impressive results on an image-recipe retrieval task. Moreover, we demonstrate that regularization via the addition of a high-level classification objective both improves retrieval performance to rival that of humans and enables semantic vector arithmetic. We postulate that these embeddings will provide a basis for further exploration of the Recipe1M+ dataset and food and cooking in general. Code, data and models are publicly available.

The Microsoft Kinect camera is becoming increasingly popular in many areas aside from entertainment, including human activity monitoring and rehabilitation. Many people, however, fail to consider the reliability and accuracy of the Kinect human pose estimation when they depend on it as a measuring system. In this paper we compare the Kinect pose estimation (skeletonization) with more established techniques for pose estimation from motion capture data, examining the accuracy of joint localization and robustness of pose estimation with respect to the orientation and occlusions. We have evaluated six physical exercises aimed at coaching of elderly population. Experimental results present pose estimation accuracy rates and corresponding error bounds for the Kinect system.

Mobile and tablet apps, ubiquitous and pervasive computing, wearable computing and domotics for health.

Microsoft Kinect camera and its skeletal tracking capabilities have been embraced by many researchers and commercial developers in various applications of real-time human movement analysis. In this paper, we evaluate the accuracy of the human kinematic motion data in the first and second generation of the Kinect system, and compare the results with an optical motion capture system. We collected motion data in 12 exercises for 10 different subjects and from three different viewpoints. We report on the accuracy of the joint localization and bone length estimation of Kinect skeletons in comparison to the motion capture. We also analyze the distribution of the joint localization offsets by fitting a mixture of Gaussian and uniform distribution models to determine the outliers in the Kinect motion data. Our analysis shows that overall Kinect 2 has more robust and more accurate tracking of human pose as compared to Kinect 1.