CVPR Daily - 2018 - Thursday

Raghu’s picks of the day: Raghu B. Krishnamurthy is a computer vision research engineer at Eyeris . He studied at the Rochester Institute of Technology, New York. “Do you want to learn more? Please stop by the Eyeris booth (#209) and we’ll be happy to give you a real-time demo.. For today, Thursday 21 2 Raghu’s Picks Thursday Oral: 14:50 (O3-3C) [H17] Feature Space Transfer for Data Augmentation Spotlights: 09:14 (S3-1B) [C14] Through-Wall Human Pose Estimation Using Radio Signals 09:14 (S3-1C) [D16] Learning to Adapt Structured Output Space for Semantic Segmentation 13:34 (S3-2A) [A10] Detecting and Recognizing Human-Object Interactions 13:34 (S3-2B) [C14] “Learning-Compression” Algorithms for Neural Net Poster: [H4] What Have We Learned From Deep Representations for Action Recognition? We do human behaviour understanding AI inside autonomous and highly automated vehicles. That includes the entire suite of face analytics, including the seven universal emotions. In addition, that includes upper- body tracking from standard 2D cameras. From that, we derive action recognition and activity prediction and we close the loop with object detection in this last consumer space or physical space of the interior of the vehicle.