Saved in:
Bibliographic Details
Main Authors: Du, Siqi, Wang, Weixi, Guo, Renzhong, Wang, Ruisheng, Tian, Yibin, Tang, Shengjun
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2309.14065
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916209044226048
author Du, Siqi
Wang, Weixi
Guo, Renzhong
Wang, Ruisheng
Tian, Yibin
Tang, Shengjun
author_facet Du, Siqi
Wang, Weixi
Guo, Renzhong
Wang, Ruisheng
Tian, Yibin
Tang, Shengjun
contents Understanding indoor scenes is crucial for urban studies. Considering the dynamic nature of indoor environments, effective semantic segmentation requires both real-time operation and high accuracy.To address this, we propose AsymFormer, a novel network that improves real-time semantic segmentation accuracy using RGB-D multi-modal information without substantially increasing network complexity. AsymFormer uses an asymmetrical backbone for multimodal feature extraction, reducing redundant parameters by optimizing computational resource distribution. To fuse asymmetric multimodal features, a Local Attention-Guided Feature Selection (LAFS) module is used to selectively fuse features from different modalities by leveraging their dependencies. Subsequently, a Cross-Modal Attention-Guided Feature Correlation Embedding (CMA) module is introduced to further extract cross-modal representations. The AsymFormer demonstrates competitive results with 54.1% mIoU on NYUv2 and 49.1% mIoU on SUNRGBD. Notably, AsymFormer achieves an inference speed of 65 FPS (79 FPS after implementing mixed precision quantization) on RTX3090, demonstrating that AsymFormer can strike a balance between high accuracy and efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2309_14065
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle AsymFormer: Asymmetrical Cross-Modal Representation Learning for Mobile Platform Real-Time RGB-D Semantic Segmentation
Du, Siqi
Wang, Weixi
Guo, Renzhong
Wang, Ruisheng
Tian, Yibin
Tang, Shengjun
Computer Vision and Pattern Recognition
Understanding indoor scenes is crucial for urban studies. Considering the dynamic nature of indoor environments, effective semantic segmentation requires both real-time operation and high accuracy.To address this, we propose AsymFormer, a novel network that improves real-time semantic segmentation accuracy using RGB-D multi-modal information without substantially increasing network complexity. AsymFormer uses an asymmetrical backbone for multimodal feature extraction, reducing redundant parameters by optimizing computational resource distribution. To fuse asymmetric multimodal features, a Local Attention-Guided Feature Selection (LAFS) module is used to selectively fuse features from different modalities by leveraging their dependencies. Subsequently, a Cross-Modal Attention-Guided Feature Correlation Embedding (CMA) module is introduced to further extract cross-modal representations. The AsymFormer demonstrates competitive results with 54.1% mIoU on NYUv2 and 49.1% mIoU on SUNRGBD. Notably, AsymFormer achieves an inference speed of 65 FPS (79 FPS after implementing mixed precision quantization) on RTX3090, demonstrating that AsymFormer can strike a balance between high accuracy and efficiency.
title AsymFormer: Asymmetrical Cross-Modal Representation Learning for Mobile Platform Real-Time RGB-D Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2309.14065