Unsupervised Disentanglement of Pose, Appearance and Background from Images and Videos
- 29 January 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. PP (99) , 1
Abstract
Unsupervised landmark learning is the task of learning semantic keypoint-like representations without the use of expensive input keypoint-level annotations. A popular approach is to factorize an image into a pose and appearance data stream, then to reconstruct the image from the factorized components. The pose representation should capture a set of consistent and tightly localized landmarks in order to facilitate reconstruction of the input image. Ultimately, we wish for our learned landmarks to focus on the foreground object of interest. However, the reconstruction task of the entire image forces the model to allocate landmarks to model the background. Using a motion-based foreground assumption, this work explores the effects of factorizing the reconstruction task into separate foreground and background reconstructions in an unsupervised way, allowing the model to condition only the foreground reconstruction on the unsupervised landmarks. Our experiments demonstrate that the proposed factorization results in landmarks that are focused on the foreground object of interest when measured against ground-truth foreground masks. Furthermore, the rendered background quality is also improved as ill-suited landmarks are no longer forced to model this content. We demonstrate this improvement via improved image fidelity in a video-prediction task. Code is available at https://github.com/NVIDIA/UnsupervisedLandmarkLearningKeywords
All Related Versions
This publication has 23 references indexed in Scilit:
- Temporal Generative Adversarial Nets with Singular Value Clipping2017
- Stacked Hourglass Networks for Human Pose Estimation2016
- WarpNet: Weakly Supervised Matching for Single-View Reconstruction2016
- Deep Learning Face Attributes in the Wild2015
- Flowing ConvNets for Human Pose Estimation in Videos2015
- Dense Optical Flow Prediction from a Static Image2015
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015
- Human3.6M: Large Scale Datasets and Predictive Methods for 3D Human Sensing in Natural EnvironmentsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2013
- Domain Adaptation for Upper Body Pose Tracking in Signed TV Broadcasts2013
- Recognizing human actions: a local SVM approach2004