Leveraging Semi-Supervised Learning in Video Sequences for Urban Scene Segmentation

@inproceedings{Chen2020LeveragingSL,
  title={Leveraging Semi-Supervised Learning in Video Sequences for Urban Scene Segmentation},
  author={Liang-Chieh Chen and Raphael Gontijo Lopes and Bowen Cheng and Maxwell D. Collins and Ekin Dogus Cubuk and Barret Zoph and Hartwig Adam and Jonathon Shlens},
  booktitle={ECCV},
  year={2020}
}
Supervised learning in large discriminative models is a mainstay for modern computer vision. Such an approach necessitates investing in large-scale human-annotated datasets for achieving state-of-the-art results. In turn, the efficacy of supervised learning may be limited by the size of the human annotated dataset. This limitation is particularly notable for image segmentation tasks, where the expense of human annotation is especially large, yet large amounts of unlabeled data may exist. In… Expand
A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentation
TLDR
It is demonstrated that the devil is in the details: a set of simple design and training techniques can collectively improve the performance of semisupervised semantic segmentation significantly. Expand
Semi-supervised Domain Adaptation for Semantic Segmentation
TLDR
This paper proposes a novel and effective two-step semi-supervised dual-domain adaptation (SSDDA) approach to address both crossand intra-domain gaps in semantic segmentation and demonstrates that the proposed approach outperforms state-of-the-art methods on two common synthetic-to-real semantic segmentsation benchmarks. Expand
A baseline for semi-supervised learning of efficient semantic segmentation models
TLDR
This work presents a simple and effective baseline which works very well both on standard and efficient architectures and shows advantage of perturbing only the student branch and presents a plausible explanation of such behaviour. Expand
PseudoSeg: Designing Pseudo Labels for Semantic Segmentation
TLDR
This work presents a simple and novel re-design of pseudo-labeling to generate well-calibrated structured pseudo labels for training with unlabeled or weakly-labeled data and demonstrates the effectiveness of the proposed pseudo- labeling strategy in both low-data and high-data regimes. Expand
Video Instance Segmentation by Instance Flow Assembly
TLDR
This work proposes a bottom-up framework equipped with a temporal context fusion module to better encode inter-frame correlations and introduces an instancelevel correspondence across adjacent frames, which is represented by a center-to-center flow, termed as instance flow, to assemble messy dense temporal correspondences. Expand
ViP-DeepLab: Learning Visual Perception with Depth-aware Video Panoptic Segmentation
In this paper, we present ViP-DeepLab, a unified model attempting to tackle the long-standing and challenging inverse projection problem in vision, which we model as restoring the point clouds fromExpand
Scaling Wide Residual Networks for Panoptic Segmentation
TLDR
This work revisits the Wide Residual Networks architecture design for the recent challenging panoptic segmentation task, which aims to unify semantic segmentation and instance segmentation in both the fast model regime and strong model regime. Expand
A Closer Look at Self-training for Zero-Label Semantic Segmentation
TLDR
This approach achieves the new state-of-the-art on PascalVOC12 and COCO-stuff datasets in the challenging generalized zero-label semantic segmentation setting, surpassing other existing methods addressing this task with more complex strategies. Expand
A Survey on Deep Learning Technique for Video Segmentation
TLDR
This survey comprehensively review two basic lines of research in this area, i.e., generic object segmentation in videos and video semantic segmentation, by introducing their respective task settings, background concepts, perceived need, development history, and main challenges. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 107 REFERENCES
Seamless Scene Segmentation
TLDR
This work introduces a novel, CNN-based architecture that can be trained end-to-end to deliver seamless scene segmentation results by means of a panoptic output format, going beyond the simple combination of independently trained segmentation and detection models. Expand
Deformable convolutional networks – COCO detection and segmentation challenge 2017 entry
  • ICCV COCO Challenge Workshop
  • 2017
The Cityscapes Dataset for Semantic Urban Scene Understanding
TLDR
This work introduces Cityscapes, a benchmark suite and large-scale dataset to train and test approaches for pixel-level and instance-level semantic labeling, and exceeds previous attempts in terms of dataset size, annotation richness, scene variability, and complexity. Expand
Object-Contextual Representations for Semantic Segmentation
TLDR
This paper addresses the semantic segmentation problem with a focus on the context aggregation strategy, and presents a simple yet effective approach, object-contextual representations, characterizing a pixel by exploiting the representation of the corresponding object class. Expand
PolyTransform: Deep Polygon Transformer for Instance Segmentation
In this paper, we propose PolyTransform, a novel instance segmentation algorithm that produces precise, geometry-preserving masks by combining the strengths of prevailing segmentation approaches andExpand
Improving Semantic Segmentation via Video Propagation and Label Relaxation
TLDR
This paper presents a video prediction-based methodology to scale up training sets by synthesizing new training samples in order to improve the accuracy of semantic segmentation networks, and introduces a novel boundary label relaxation technique that makes training robust to annotation noise and propagation artifacts along object boundaries. Expand
Semantic Video Segmentation by Gated Recurrent Flow Propagation
TLDR
A deep, end-to-end trainable methodology for video segmentation that is capable of leveraging the information present in unlabeled data, besides sparsely labeled frames, in order to improve semantic estimates. Expand
Understanding Convolution for Semantic Segmentation
TLDR
DUC is designed to generate pixel-level prediction, which is able to capture and decode more detailed information that is generally missing in bilinear upsampling, and a hybrid dilated convolution (HDC) framework in the encoding phase is proposed. Expand
Semantic Video CNNs Through Representation Warping
TLDR
A key insight of this work is that fast optical flow methods can be combined with many different CNN architectures for improved performance and end-to-end training. Expand
Weakly-and semisupervised learning of a deep convolutional network for semantic image segmentation
  • ICCV
  • 2015
...
1
2
3
4
5
...