How to Make Cross Encoder a Good Teacher for Efficient Image-Text Retrieval?

• URL: http://arxiv.org/abs/2407.07479v1
• Date: Wed, 10 Jul 2024 09:10:01 GMT
• Title: How to Make Cross Encoder a Good Teacher for Efficient Image-Text Retrieval?
• Authors: Yuxin Chen, Zongyang Ma, Ziqi Zhang, Zhongang Qi, Chunfeng Yuan, Bing Li, Junfu Pu, Ying Shan, Xiaojuan Qi, Weiming Hu,
• Abstract summary: Cross-modal similarity score distribution of cross-encoder is more concentrated while the result of dual-encoder is nearly normal. Only the relative order between hard negatives conveys valid knowledge while the order information between easy negatives has little significance. We propose a novel Contrastive Partial Ranking Distillation (DCPR) method which implements the objective of mimicking relative order between hard negative samples with contrastive learning.
• Score: 99.87554379608224
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Dominant dual-encoder models enable efficient image-text retrieval but suffer from limited accuracy while the cross-encoder models offer higher accuracy at the expense of efficiency. Distilling cross-modality matching knowledge from cross-encoder to dual-encoder provides a natural approach to harness their strengths. Thus we investigate the following valuable question: how to make cross-encoder a good teacher for dual-encoder? Our findings are threefold:(1) Cross-modal similarity score distribution of cross-encoder is more concentrated while the result of dual-encoder is nearly normal making vanilla logit distillation less effective. However ranking distillation remains practical as it is not affected by the score distribution.(2) Only the relative order between hard negatives conveys valid knowledge while the order information between easy negatives has little significance.(3) Maintaining the coordination between distillation loss and dual-encoder training loss is beneficial for knowledge transfer. Based on these findings we propose a novel Contrastive Partial Ranking Distillation (CPRD) method which implements the objective of mimicking relative order between hard negative samples with contrastive learning. This approach coordinates with the training of the dual-encoder effectively transferring valid knowledge from the cross-encoder to the dual-encoder. Extensive experiments on image-text retrieval and ranking tasks show that our method surpasses other distillation methods and significantly improves the accuracy of dual-encoder.

Related papers

• Triple-View Knowledge Distillation for Semi-Supervised Semantic Segmentation [54.23510028456082]
  We propose a Triple-view Knowledge Distillation framework, termed TriKD, for semi-supervised semantic segmentation. The framework includes the triple-view encoder and the dual-frequency decoder.
  arXiv  Detail & Related papers  (2023-09-22T01:02:21Z)
• Efficient Nearest Neighbor Search for Cross-Encoder Models using Matrix Factorization [60.91600465922932]
  We present an approach that avoids the use of a dual-encoder for retrieval, relying solely on the cross-encoder. Our approach provides test-time recall-vs-computational cost trade-offs superior to the current widely-used methods.
  arXiv  Detail & Related papers  (2022-10-23T00:32:04Z)
• Revisiting Code Search in a Two-Stage Paradigm [67.02322603435628]
  TOSS is a two-stage fusion code search framework. It first uses IR-based and bi-encoder models to efficiently recall a small number of top-k code candidates. It then uses fine-grained cross-encoders for finer ranking.
  arXiv  Detail & Related papers  (2022-08-24T02:34:27Z)
• AWEncoder: Adversarial Watermarking Pre-trained Encoders in Contrastive Learning [18.90841192412555]
  We introduce AWEncoder, an adversarial method for watermarking the pre-trained encoder in contrastive learning. The proposed work enjoys pretty good effectiveness and robustness on different contrastive learning algorithms and downstream tasks.
  arXiv  Detail & Related papers  (2022-08-08T07:23:37Z)
• LoopITR: Combining Dual and Cross Encoder Architectures for Image-Text Retrieval [117.15862403330121]
  We propose LoopITR, which combines dual encoders and cross encoders in the same network for joint learning. Specifically, we let the dual encoder provide hard negatives to the cross encoder, and use the more discriminative cross encoder to distill its predictions back to the dual encoder.
  arXiv  Detail & Related papers  (2022-03-10T16:41:12Z)
• Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations [22.40667024030858]
  Bi-encoders produce fixed-dimensional sentence representations and are computationally efficient. Cross-encoders can leverage their attention heads to exploit inter-sentence interactions for better performance. Trans-Encoder combines the two learning paradigms into an iterative joint framework to simultaneously learn enhanced bi- and cross-encoders.
  arXiv  Detail & Related papers  (2021-09-27T14:06:47Z)
• CoDERT: Distilling Encoder Representations with Co-learning for Transducer-based Speech Recognition [14.07385381963374]
  We show that the transducer's encoder outputs naturally have a high entropy and contain rich information about acoustically similar word-piece confusions. We introduce an auxiliary loss to distill the encoder logits from a teacher transducer's encoder, and explore training strategies where this encoder distillation works effectively.
  arXiv  Detail & Related papers  (2021-06-14T20:03:57Z)
• Two are Better than One: Joint Entity and Relation Extraction with Table-Sequence Encoders [13.999110725631672]
  Two different encoders are designed to help each other in the representation learning process. Our experiments confirm the advantages of having em two encoders over em one encoder.
  arXiv  Detail & Related papers  (2020-10-08T09:10:55Z)
• Auto-Encoding Twin-Bottleneck Hashing [141.5378966676885]
  This paper proposes an efficient and adaptive code-driven graph. It is updated by decoding in the context of an auto-encoder. Experiments on benchmarked datasets clearly show the superiority of our framework over the state-of-the-art hashing methods.
  arXiv  Detail & Related papers  (2020-02-27T05:58:12Z)

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.