ECCV 2018 Daily - Tuesday

Daily Tuesday 11 Sarah’s picks of the day: P-2A-25 Space-time Knowledge for Unpaired Image-to-Image Translation P-2A-62 Reinforced Temporal Attention and Split-Rate Transfer for … P-2A-88 Adversarial Geometry-Aware Human Motion Prediction P-2B-36 Image Super-Resolution Using Very Deep Residual Channel Attention Networks P-2B-75 BodyNet: Volumetric Inference of 3D Human Body Shapes Sarah Ostadabbas is Assistant Professor of Electrical and Computer Engineering at the Northeastern University in Boston, USA. She is the Director of the Augmented Cognition Lab (ACLab) , which works at the intersection of computer vision and machine learning with the goal of understanding, detecting, and predicting human behaviors through a series of physical, psychological, and cognitive state estimation. For a robust and efficient state estimation, she represents the state of the world in a low-dimensional subspace, called "pose", which is a succinct interpretable representation of the important information in the state. The ACLab uses machine intelligence to solve these pose estimation problems, including estimating poses from human faces and bodies as well as poses of environments and scenes. Sarah’s Picks For today, Tuesday 11 2 Sarah invites us to their paper's live demo which will take place today at 10am during Live Demo session 2A (#3). In this demo, Shuangjun Liu (PhD student at ACLab) has implemented his reposing model in real time, such that you can arbitrarily move in front of our webcam and a picture of your choice will mimic your pose in real time, while preserving the inner space of the target image! Sarah’s lab ACLab attended ECCV2018 with 3 presentations: • A Semi-Supervised Data Augmentation Approach using 3D Graphical Engines , Poster at the 9th International HBU ECCV Workshop • Inner Space Preserving Generative Pose Machine , Poster at the ECCV Main Track • Inner Space Preserving Generative Pose Machine (ISP-GPM) , Live Demo at ECCV Sarah Ostadabbas and her PhD student, Shuangjun Liu. Shuangjun could not travel to the ECCV2018 since he is on a single-entry student visa. Sarah was featured as Woman in Computer Vision on Computer Vision News of December 2017 .