Cost Aggregation

Cost aggregation is a crucial process in computer vision, aiming to improve the accuracy and efficiency of matching similar features or pixels across images, particularly in tasks like stereo matching and multi-view stereo. Recent research emphasizes developing novel aggregation methods using convolutional neural networks (CNNs) and transformers, focusing on improving geometric consistency, handling high-dimensional cost volumes efficiently, and integrating feature and cost aggregation for enhanced performance. These advancements lead to significant improvements in accuracy and speed for various applications, including depth estimation, semantic correspondence, and few-shot segmentation, enabling more robust and efficient computer vision systems.

Papers

June 28, 2024

LightStereo: Channel Boost Is All Your Need for Efficient 2D Cost Aggregation
Xianda Guo, Chenming Zhang, Youmin Zhang, Wenzhao Zheng, Dujun Nie, Matteo Poggi, Long Chen
Community Need Stereo Matching Scene Datasets Stereo Matching Network Cost Aggregation Channel Importance

April 11, 2024

GoMVS: Geometrically Consistent Cost Aggregation for Multi-View Stereo
Jiang Wu, Rui Li, Haofei Xu, Wenxun Zhao, Yu Zhu, Jinqiu Sun, Yanning Zhang
Multi View Stereo Geometric Consistency Hypothesis Space Cost Aggregation Hierarchical Propagation

March 17, 2024

Unifying Feature and Cost Aggregation with Transformers for Semantic and Visual Correspondence
Sunghwan Hong, Seokju Cho, Seungryong Kim, Stephen Lin
Transformer Megatron Decepticons Semantic Description Feature Aggregation Visual Correspondence Dense Matching Unified Data Cost Aggregation

January 22, 2024

Boosting Multi-view Stereo with Late Cost Aggregation
Jiang Wu, Rui Li, Yu Zhu, Wenxun Zhao, Jinqiu Sun, Yanning Zhang
Multi View Stereo Depth Network Pairwise Distance Learning Based Multi View Stereo Temporal Aggregation Cost Aggregation

June 13, 2023

Continuous Cost Aggregation for Dual-Pixel Disparity Extraction
Sagi Monin, Sagi Katz, Georgios Evangelidis
Stereo Image Disparity Estimation Stereo Matching Algorithm Cost Aggregation Dual Pixel

May 17, 2023

CostFormer:Cost Transformer for Cost Aggregation in Multi-view Stereo
Weitao Chen, Hongbin Xu, Zhipeng Zhou, Yang Liu, Baigui Sun, Wenxiong Kang, Xuansong Xie
Transformer Based Multi View Multi View Stereo Cost Aggregation Dense Residual Connected Transformer

September 19, 2022

Integrative Feature and Cost Aggregation with Transformers for Dense Correspondence
Sunghwan Hong, Seokju Cho, Seungryong Kim, Stephen Lin
Transformer Megatron Decepticons Self Attention Layer Dense Correspondence Transformer Based Neural Network Dense Matching Cost Volume Cost Aggregation

July 22, 2022

Cost Aggregation with 4D Convolutional Swin Transformer for Few-Shot Segmentation
Sunghwan Hong, Seokju Cho, Jisu Nam, Stephen Lin, Seungryong Kim
Segmentation Based Approach Swin Transformer Shot Segmentation Cost Aggregation Correlation Map Convolutional Inductive Bias

April 5, 2022

Joint Learning of Feature Extraction and Cost Aggregation for Semantic Correspondence
Jiwon Kim, Youngjo Min, Mira Kim, Seungryong Kim
Feature Extraction Weakly Supervised Joint Learning Dense Correspondence Semantic Correspondence Cost Aggregation Confidence Aware Contrastive

February 14, 2022

CATs++: Boosting Cost Aggregation with Convolutions and Transformers
Seokju Cho, Sunghwan Hong, Seungryong Kim
Convolutional Neural Network Transformer Megatron Decepticons Direct Convolution Receptive Field Image Matching Cost Aggregation

December 22, 2021

Cost Aggregation Is All You Need for Few-Shot Segmentation
Sunghwan Hong, Seokju Cho, Jisu Nam, Seungryong Kim
Segmentation Based Approach Shot Segmentation Aggregation Function Cost Aggregation Convolutional Inductive Bias