PrevMatch: Revisiting and Maximizing Temporal Knowledge in Semi-Supervised Semantic Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shin, Wooseok, Park, Hyun Joon, Kim, Jin Sob, Yun, Juan, Park, Se Hong, Han, Sung Won
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918272440467456
author Shin, Wooseok
Park, Hyun Joon
Kim, Jin Sob
Yun, Juan
Park, Se Hong
Han, Sung Won
author_facet Shin, Wooseok
Park, Hyun Joon
Kim, Jin Sob
Yun, Juan
Park, Se Hong
Han, Sung Won
contents In semi-supervised semantic segmentation, the Mean Teacher- and co-training-based approaches are employed to mitigate confirmation bias and coupling problems. However, despite their high performance, these approaches frequently involve complex training pipelines and a substantial computational burden, limiting the scalability and compatibility of these methods. In this paper, we propose a PrevMatch framework that effectively mitigates the aforementioned limitations by maximizing the utilization of the temporal knowledge obtained during the training process. The PrevMatch framework relies on two core strategies: (1) we reconsider the use of temporal knowledge and thus directly utilize previous models obtained during training to generate additional pseudo-label guidance, referred to as previous guidance. (2) we design a highly randomized ensemble strategy to maximize the effectiveness of the previous guidance. PrevMatch, a simple yet effective plug-in method, can be seamlessly integrated into existing semi-supervised learning frameworks with minimal computational overhead. Experimental results on three benchmark semantic segmentation datasets show that incorporating PrevMatch into existing methods significantly improves their performance. Furthermore, our analysis indicates that PrevMatch facilitates stable optimization during training, resulting in improved generalization performance.
format Preprint
id arxiv_https___arxiv_org_abs_2405_20610
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PrevMatch: Revisiting and Maximizing Temporal Knowledge in Semi-Supervised Semantic Segmentation
Shin, Wooseok
Park, Hyun Joon
Kim, Jin Sob
Yun, Juan
Park, Se Hong
Han, Sung Won
Computer Vision and Pattern Recognition
I.4.6
In semi-supervised semantic segmentation, the Mean Teacher- and co-training-based approaches are employed to mitigate confirmation bias and coupling problems. However, despite their high performance, these approaches frequently involve complex training pipelines and a substantial computational burden, limiting the scalability and compatibility of these methods. In this paper, we propose a PrevMatch framework that effectively mitigates the aforementioned limitations by maximizing the utilization of the temporal knowledge obtained during the training process. The PrevMatch framework relies on two core strategies: (1) we reconsider the use of temporal knowledge and thus directly utilize previous models obtained during training to generate additional pseudo-label guidance, referred to as previous guidance. (2) we design a highly randomized ensemble strategy to maximize the effectiveness of the previous guidance. PrevMatch, a simple yet effective plug-in method, can be seamlessly integrated into existing semi-supervised learning frameworks with minimal computational overhead. Experimental results on three benchmark semantic segmentation datasets show that incorporating PrevMatch into existing methods significantly improves their performance. Furthermore, our analysis indicates that PrevMatch facilitates stable optimization during training, resulting in improved generalization performance.
title PrevMatch: Revisiting and Maximizing Temporal Knowledge in Semi-Supervised Semantic Segmentation
topic Computer Vision and Pattern Recognition
I.4.6
url https://arxiv.org/abs/2405.20610