Learning Semantic Segmentation from Multiple Datasets with Label Shifts

• URL: http://arxiv.org/abs/2202.14030v1
• Date: Mon, 28 Feb 2022 18:55:19 GMT
• Title: Learning Semantic Segmentation from Multiple Datasets with Label Shifts
• Authors: Dongwan Kim, Yi-Hsuan Tsai, Yumin Suh, Masoud Faraki, Sparsh Garg, Manmohan Chandraker, Bohyung Han
• Abstract summary: This paper proposes UniSeg, an effective approach to automatically train models across multiple datasets with differing label spaces. Specifically, we propose two losses that account for conflicting and co-occurring labels to achieve better generalization performance in unseen domains.
• Score: 101.24334184653355
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: With increasing applications of semantic segmentation, numerous datasets have been proposed in the past few years. Yet labeling remains expensive, thus, it is desirable to jointly train models across aggregations of datasets to enhance data volume and diversity. However, label spaces differ across datasets and may even be in conflict with one another. This paper proposes UniSeg, an effective approach to automatically train models across multiple datasets with differing label spaces, without any manual relabeling efforts. Specifically, we propose two losses that account for conflicting and co-occurring labels to achieve better generalization performance in unseen domains. First, a gradient conflict in training due to mismatched label spaces is identified and a class-independent binary cross-entropy loss is proposed to alleviate such label conflicts. Second, a loss function that considers class-relationships across datasets is proposed for a better multi-dataset training scheme. Extensive quantitative and qualitative analyses on road-scene datasets show that UniSeg improves over multi-dataset baselines, especially on unseen datasets, e.g., achieving more than 8% gain in IoU on KITTI averaged over all the settings.

Related papers

• Continuous Contrastive Learning for Long-Tailed Semi-Supervised Recognition [50.61991746981703]
  Current state-of-the-art LTSSL approaches rely on high-quality pseudo-labels for large-scale unlabeled data. This paper introduces a novel probabilistic framework that unifies various recent proposals in long-tail learning. We introduce a continuous contrastive learning method, CCL, extending our framework to unlabeled data using reliable and smoothed pseudo-labels.
  arXiv  Detail & Related papers  (2024-10-08T15:06:10Z)
• Automated Label Unification for Multi-Dataset Semantic Segmentation with GNNs [48.406728896785296]
  We propose a novel approach to automatically construct a unified label space across multiple datasets using graph neural networks. Unlike existing methods, our approach facilitates seamless training without the need for additional manual reannotation or taxonomy reconciliation.
  arXiv  Detail & Related papers  (2024-07-15T08:42:10Z)
• Tackling Diverse Minorities in Imbalanced Classification [80.78227787608714]
  Imbalanced datasets are commonly observed in various real-world applications, presenting significant challenges in training classifiers. We propose generating synthetic samples iteratively by mixing data samples from both minority and majority classes. We demonstrate the effectiveness of our proposed framework through extensive experiments conducted on seven publicly available benchmark datasets.
  arXiv  Detail & Related papers  (2023-08-28T18:48:34Z)
• TransPOS: Transformers for Consolidating Different POS Tagset Datasets [0.8399688944263843]
  This paper considers two datasets that label part-of-speech (POS) tags under different tagging schemes. It proposes a novel supervised architecture employing Transformers to tackle the problem of consolidating two completely disjoint datasets.
  arXiv  Detail & Related papers  (2022-09-24T08:43:53Z)
• Cross-Domain Few-Shot Classification via Inter-Source Stylization [11.008292768447614]
  Cross-Domain Few-Shot Classification (CDFSC) is to accurately classify a target dataset with limited labelled data. This paper proposes a solution that makes use of multiple source domains without the need for additional labeling costs.
  arXiv  Detail & Related papers  (2022-08-17T01:44:32Z)
• Multi-domain semantic segmentation with overlapping labels [1.4120796122384087]
  We propose a principled method for seamless learning on datasets with overlapping classes based on partial labels and probabilistic loss. Our method achieves competitive within-dataset and cross-dataset generalization, as well as ability to learn visual concepts which are not separately labeled in any of the training datasets.
  arXiv  Detail & Related papers  (2021-08-25T13:25:41Z)
• Multi-dataset Pretraining: A Unified Model for Semantic Segmentation [97.61605021985062]
  We propose a unified framework, termed as Multi-Dataset Pretraining, to take full advantage of the fragmented annotations of different datasets. This is achieved by first pretraining the network via the proposed pixel-to-prototype contrastive loss over multiple datasets. In order to better model the relationship among images and classes from different datasets, we extend the pixel level embeddings via cross dataset mixing.
  arXiv  Detail & Related papers  (2021-06-08T06:13:11Z)
• DAIL: Dataset-Aware and Invariant Learning for Face Recognition [67.4903809903022]
  To achieve good performance in face recognition, a large scale training dataset is usually required. It is problematic and troublesome to naively combine different datasets due to two major issues. Naively treating the same person as different classes in different datasets during training will affect back-propagation. manually cleaning labels may take formidable human efforts, especially when there are millions of images and thousands of identities.
  arXiv  Detail & Related papers  (2021-01-14T01:59:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.