PARS: Pseudo-Label Aware Robust Sample Selection for Learning with Noisy Labels

Goel, Arushi; Jiao, Yunlong; Massiah, Jordan

doi:10.48550/arxiv.2201.10836

Cited by 2 publications

(2 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A major issue for ST approaches is confirmation bias, where the student model would accumulate errors from the teacher model when learning with inaccurate pseudo-labels (e.g. Goel et al, 2022;.…”

Section: Related Workmentioning

confidence: 99%

Rethinking Semi-supervised Learning with Language Models

Zhengxiang¹,

Tonolini²,

Αλέτρας³

et al. 2023

Findings of the Association for Computational Linguistics: ACL 2023

View full text Add to dashboard Cite

Semi-supervised learning (SSL) is a popular setting aiming to effectively utilize unlabelled data to improve model performance in downstream natural language processing (NLP) tasks. Currently, there are two popular approaches to make use of unlabelled data: Self-training (ST) and Task-adaptive pre-training (TAPT). ST uses a teacher model to assign pseudo-labels to the unlabelled data, while TAPT continues pre-training on the unlabelled data before finetuning. To the best of our knowledge, the effectiveness of TAPT in SSL tasks has not been systematically studied, and no previous work has directly compared TAPT and ST in terms of their ability to utilize the pool of unlabelled data. In this paper, we provide an extensive empirical study comparing five state-of-the-art ST approaches and TAPT across various NLP tasks and data sizes, including in-and out-of-domain settings. Surprisingly, we find that TAPT is a strong and more robust SSL learner, even when using just a few hundred unlabelled samples or in the presence of domain shifts, compared to more sophisticated ST approaches, and tends to bring greater improvements in SSL than in fully-supervised settings. Our further analysis demonstrates the risks of using ST approaches when the size of labelled or unlabelled data is small or when domain shifts exist. We offer a fresh perspective for future SSL research, suggesting the use of unsupervised pre-training objectives over dependency on pseudo labels. 1 * This work was done during an internship at Amazon, Alexa Shopping.

show abstract

Section: Related Workmentioning

confidence: 99%

Rethinking Semi-supervised Learning with Language Models

Zhengxiang¹,

Tonolini²,

Αλέτρας³

et al. 2023

Findings of the Association for Computational Linguistics: ACL 2023

View full text Add to dashboard Cite

show abstract

“…Employing object detection methods in SSOD poses several potential challenges that must be carefully dealt with to obtain reasonable performance. These factors include overfitting of the labeled data [39], pseudo-label noise [11], bias induced through label imbalance [18,33], and poor detection performance on small objects [58]. Recently, DETRbased [2,17,19,26,40,57,61] SSOD methods [48,58] remove the need for traditional components like NMS.…”

Section: Introductionmentioning

confidence: 99%

Mask-Aware Semi-Supervised Object Detection in Floor Plans

Shehzadi¹,

Hashmi²,

Pagani³

et al. 2022

Preprint

View full text Add to dashboard Cite

Research has been growing on object detection using semi-supervised methods in past few years. We examine the intersection of these two areas for floor-plan objects to promote the research objective of detecting more accurate objects with less labelled data. The floor-plan objects include different furniture items with multiple types of the same class, and this high inter-class similarity impacts the performance of prior methods. In this paper, we present Mask R-CNN based semi-supervised approach that provides pixel-to-pixel alignment to generate individual annotation masks for each class to mine the inter-class similarity. The semi-supervised approach has a student-teacher network that pulls information from the teacher network and feeds it to the student network. The teacher network uses unlabeled data to form pseudo-boxes, and the student network uses both unlabeled data with the pseudo boxes and labelled data as ground truth for training. It learns representations of furniture items by combining labelled and unlabeled data. On the Mask R-CNN detector with ResNet-101 backbone network, the proposed approach achieves mAP of 98.8%, 99.7%, and 99.8% with only 1%, 5% and 10% labelled data, respectively. Our experiment affirms the efficiency of the proposed approach as it outperforms the fully supervised counterpart using only 10% of the labels.

show abstract

PARS: Pseudo-Label Aware Robust Sample Selection for Learning with Noisy Labels

Cited by 2 publications

References 21 publications

Rethinking Semi-supervised Learning with Language Models

Rethinking Semi-supervised Learning with Language Models

Mask-Aware Semi-Supervised Object Detection in Floor Plans

Contact Info

Product

Resources

About