Skip to main navigation Skip to search Skip to main content

Transpose and Mask: Simple and Effective Logit-Based Knowledge Distillation for Multi-attribute and Multi-label Classification

  • Beihang University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Knowledge distillation (KD) improves a student network by transferring knowledge from a teacher network. Although KD has been extensively studied in single-labeled image classification, it is not well explored under the scope of multi-attribute and multi-label classification. We observe that the logit-based KD method for the single-label scene utilizes information from multiple classes in a single sample, but we find such logits are less informative in the multi-label scene. To address this challenge in the multi-label scene, we design a Transpose method to extract information from multiple samples in a batch instead of a single sample. We further note that certain classes may lack positive samples in a batch, which can negatively impact the training process. To address this issue, we design another strategy, the Mask, to prevent the influence of negative samples. To conclude, we propose Transpose and Mask Knowledge Distillation (TM-KD), a simple and effective logit-based KD framework for multi-attribute and multi-label classification. The effectiveness of TM-KD is confirmed by experiments on multiple tasks and datasets, including pedestrian attribute recognition (PETA, PETA-zs, PA100k), clothing attribute recognition (Clothing Attributes Dataset), and multi-label classification (MS COCO), showing impressive and consistent performance gains.

Original languageEnglish
Title of host publicationPattern Recognition and Computer Vision - 6th Chinese Conference, PRCV 2023, Proceedings
EditorsQingshan Liu, Hanzi Wang, Rongrong Ji, Zhanyu Ma, Weishi Zheng, Hongbin Zha, Xilin Chen, Liang Wang
PublisherSpringer Science and Business Media Deutschland GmbH
Pages273-284
Number of pages12
ISBN (Print)9789819985487
DOIs
StatePublished - 2024
Event6th Chinese Conference on Pattern Recognition and Computer Vision, PRCV 2023 - Xiamen, China
Duration: 13 Oct 202315 Oct 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14434 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference6th Chinese Conference on Pattern Recognition and Computer Vision, PRCV 2023
Country/TerritoryChina
CityXiamen
Period13/10/2315/10/23

Keywords

  • Knowledge distillation
  • multi-attribute classification
  • multi-label classification

Fingerprint

Dive into the research topics of 'Transpose and Mask: Simple and Effective Logit-Based Knowledge Distillation for Multi-attribute and Multi-label Classification'. Together they form a unique fingerprint.

Cite this