RECORD
Transductive CLIP with Class-Conditional Contrastive Learning
Computer Vision and Pattern Recognition · Submitted 13 Jun 2022
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim Experiments on multiple benchmark datasets demonstrate the substantial improvements over other state-of-the-art methods.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
Inspired by the remarkable zero-shot generalization capacity of vision-language pre-trained model, we seek to leverage the supervision from CLIP model to alleviate the burden of data labeling. However, such supervision inevitably contains the label noise, which significantly degrades the discriminative power of the classification model. In this work, we propose Transductive CLIP, a novel framework for learning a classification network with noisy labels from scratch. Firstly, a class-conditional contrastive learning mechanism is proposed to mitigate the reliance on pseudo labels and boost the tolerance to noisy labels. Secondly, ensemble labels is adopted as a pseudo label updating strategy to stabilize the training of deep neural networks with noisy labels. This framework can reduce the impact of noisy labels from CLIP model effectively by combining both techniques. Experiments on multiple benchmark datasets demonstrate the substantial improvements over other state-of-the-art methods.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Inspired by the remarkable zero-shot generalization capacity of vision-language pre-trained model, we seek to leverage the supervision from CLIP model to alleviate the burden of data labeling. | background 0.99 | 0.17 | 0.03 | asserted 0.74 |
| s1 | However, such supervision inevitably contains the label noise, which significantly degrades the discriminative power of the classification model. | background 0.74 | 0.13 | 0.03 | asserted 1.00 |
| s2 | In this work, we propose Transductive CLIP, a novel framework for learning a classification network with noisy labels from scratch. | contribution 0.99 | 0.83 | 0.10 | asserted 0.99 |
| s3 | Firstly, a class-conditional contrastive learning mechanism is proposed to mitigate the reliance on pseudo labels and boost the tolerance to noisy labels. | method 0.81 | 0.87 | 0.08 | asserted 0.98 |
| s4 | Secondly, ensemble labels is adopted as a pseudo label updating strategy to stabilize the training of deep neural networks with noisy labels. | method 0.93 | 0.84 | 0.05 | asserted 1.00 |
| s5 | This framework can reduce the impact of noisy labels from CLIP model effectively by combining both techniques. | result 0.34 | 0.67 | 0.33 | asserted 0.70 |
| s6 | Experiments on multiple benchmark datasets demonstrate the substantial improvements over other state-of-the-art methods. | result 1.00 | 0.94 | 0.60 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
Inspired by the remarkable zero-shot generalization capacity of vision-language pre-trained model, we seek to leverage the supervision from CLIP model to alleviate the burden of data labeling. However, such supervision inevitably contains the label noise, which significantly degrades the discriminative power of the classification model. In this work, we propose Transductive CLIP, a novel framework for learning a classification network with noisy labels from scratch. Firstly, a class-conditional contrastive learning mechanism is proposed to mitigate the reliance on pseudo labels and boost the tolerance to noisy labels. Secondly, ensemble labels is adopted as a pseudo label updating strategy to stabilize the training of deep neural networks with noisy labels. This framework can reduce the impact of noisy labels from CLIP model effectively by combining both techniques. Experiments on multiple benchmark datasets demonstrate the substantial improvements over other state-of-the-art methods.
Human review
How the literature uses it
What this paper does with the arXiv papers it cites
What it cites · 1 sentence without a confident label
Cites as background 1 paper
-
Recently, Contrastive Language-Image Pretraining (CLIP) [ref] has emerged as a promising alternative for generalizing vision tasks.
Learning Transferable Visual Models From Natural Language Supervision · Introduction · 0.94
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.