Compressing Visual-linguistic Model via Knowledge Distillation

Zhiyuan Fang, Jianfeng Wang, Xiaowei Hu, Lijuan Wang, Yezhou Yang, Zicheng Liu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

34 Scopus citations


Despite exciting progress in pre-training for visual-linguistic (VL) representations, very few aspire to a small VL model. In this paper, we study knowledge distillation (KD) to effectively compress a transformer based large VL model into a small VL model. The major challenge arises from the inconsistent regional visual tokens extracted from different detectors of Teacher and Student, resulting in the misalignment of hidden representations and attention distributions. To address the problem, we retrain and adapt the Teacher by using the same region proposals from Student's detector while the features are from Teacher's own object detector. With aligned network inputs, the adapted Teacher is capable of transferring the knowledge through the intermediate representations. Specifically, we use the mean square error loss to mimic the attention distribution inside the transformer block, and present a token-wise noise contrastive loss to align the hidden state by contrasting with negative representations stored in a sample queue. To this end, we show that our proposed distillation significantly improves the performance of small VL models on image captioning and visual question answering tasks. It reaches 120.8 in CIDEr score on COCO captioning, an improvement of 5.1 over its non-distilled counterpart; and an accuracy of 69.8 on VQA 2.0, a 0.8 gain from the baseline. Our extensive experiments and ablations confirm the effectiveness of VL distillation in both pre-training and fine-tuning stages.

Original languageEnglish (US)
Title of host publicationProceedings - 2021 IEEE/CVF International Conference on Computer Vision, ICCV 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
Number of pages11
ISBN (Electronic)9781665428125
StatePublished - 2021
Event18th IEEE/CVF International Conference on Computer Vision, ICCV 2021 - Virtual, Online, Canada
Duration: Oct 11 2021Oct 17 2021

Publication series

NameProceedings of the IEEE International Conference on Computer Vision
ISSN (Print)1550-5499


Conference18th IEEE/CVF International Conference on Computer Vision, ICCV 2021
CityVirtual, Online

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition


Dive into the research topics of 'Compressing Visual-linguistic Model via Knowledge Distillation'. Together they form a unique fingerprint.

Cite this