Small logo of ETH main building ETH Zurich : Computer Science : Pervasive Computing : Distributed Systems : Education : Student Projects : Abstract

Temporal Context in Human Pose Estimation (B)

Status: Abgeschlossen

Human pose estimation and action recognition are two closely interlinked tasks. While action recognition approaches successfully make use of temporal information, pose estimation approaches largely relies on single image prediction. In this thesis, we investigate how pose estimation can be improved by adding temporal context.

Requirements: solid C/C++ and/or Python programming skills, background in image processing is necessary. Experience with CNNs is a strong plus

Keywords: convolutional neural network, GPU, image processing

Student/Bearbeitet von: Susanne Keller
Contact/Ansprechpartner: Jie Song, Gábor Sörös

ETH ZurichDistributed Systems Group
Last updated February 3 2017 12:31:34 PM MET webvs