Leveraging Motion Priors in Videos for Improving Human Segmentation

@inproceedings{Chen2018LeveragingMP,
  title={Leveraging Motion Priors in Videos for Improving Human Segmentation},
  author={Yu-Ting Chen and Wen-Yen Chang and Hai-Lun Lu and Tingfan Wu and Min Sun},
  booktitle={ECCV},
  year={2018}
}
Despite many advances in deep-learning based semantic segmentation, performance drop due to distribution mismatch is often encountered in the real world. Recently, a few domain adaptation and active learning approaches have been proposed to mitigate the performance drop. However, very little attention has been made toward leveraging information in videos which are naturally captured in most camera systems. In this work, we propose to leverage “motion prior” in videos for improving human… Expand
2 Citations
Estimating optical flow with convolutional neural networks
TLDR
This thesis presents an orthogonal approach to estimate optical flow with Convolutional Neural Networks and shows that such networks are able to learn a better heuristic than engineered methods and are state of the art in motion boundary and occlusion estimation. Expand
Bias-Aware Heapified Policy for Active Learning
TLDR
This paper proposes a bias-aware policy network called Heapified Active Learning (HAL), which prevents overconfidence, and improves sample efficiency of policy learning by heapified structure without ignoring global inforamtion(overview of the whole unlabeled set). Expand

References

SHOWING 1-10 OF 52 REFERENCES
Unsupervised Domain Adaptation for Semantic Segmentation with GANs
TLDR
This work proposes an approach based on Generative Adversarial Networks (GANs) that brings the embeddings closer in the learned feature space and can achieve state of the art results on two challenging scenarios of synthetic to real domain adaptation. Expand
Learning from Synthetic Data: Addressing Domain Shift for Semantic Segmentation
TLDR
This work proposes an approach based on Generative Adversarial Networks (GANs) that brings the embeddings closer in the learned feature space and can achieve state of the art results on two challenging scenarios of synthetic to real domain adaptation. Expand
FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation
TLDR
This paper introduces the first domain adaptive semantic segmentation method, proposing an unsupervised adversarial approach to pixel prediction problems, and outperforms baselines across different settings on multiple large-scale datasets. Expand
SegFlow: Joint Learning for Video Object Segmentation and Optical Flow
TLDR
This paper proposes an end-to-end trainable network, SegFlow, for simultaneously predicting pixel-wise object segmentation and optical flow in videos, and demonstrates that introducing optical flow improves the performance of segmentation, against the state-of-the-art algorithms. Expand
Human action segmentation with hierarchical supervoxel consistency
TLDR
A hierarchical MRF model to bridge low-level video fragments with high-level human motion and appearance is proposed and novel higher-order potentials connect different levels of the supervoxel hierarchy to enforce the consistency of the human segmentation by pulling from different segment-scales. Expand
1000fps human segmentation with deep convolutional neural networks
TLDR
A super-fast and highly accurate human segmentation method with very deep convolutional neural networks and a comprehensive study on the proposed approach, including different net structures, various techniques of alleviating over-fitting, and performance enhancement with different extra data is provided. Expand
Segmentation of Moving Objects by Long Term Video Analysis
TLDR
This paper demonstrates that motion will be exploited most effectively, if it is regarded over larger time windows, and suggests working with a paradigm that starts with semi-dense motion cues first and that fills up textureless areas afterwards based on color. Expand
FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks
TLDR
The concept of end-to-end learning of optical flow is advanced and it work really well, and faster variants that allow optical flow computation at up to 140fps with accuracy matching the original FlowNet are presented. Expand
Curriculum Domain Adaptation for Semantic Segmentation of Urban Scenes
TLDR
This work proposes a curriculum-style learning approach to minimize the domain gap in semantic segmentation, which significantly outperforms the baselines as well as the only known existing approach to the same problem. Expand
No More Discrimination: Cross City Adaptation of Road Scene Segmenters
TLDR
This work proposes an unsupervised learning approach to adapt road scene segmenters across different cities by advancing a joint global and class-specific domain adversarial learning framework, and shows that this method improves the performance of semantic segmentation in multiple cities across continents. Expand
...
1
2
3
4
5
...