Ashish Patel 🇮🇳’s Post

𝗗𝗮𝘆-𝟰𝟳𝟮 𝗖𝗼𝗺𝗽𝘂𝘁𝗲𝗿 𝗩𝗶𝘀𝗶𝗼𝗻 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 MVSTER: Epipolar Transformer for Efficient Multi-View Stereo by Chinese Academy of Sciences Follow me for a similar post: Ashish Patel ------------------------------------------------------------------- 𝗜𝗻𝘁𝗲𝗿𝗲𝘀𝘁𝗶𝗻𝗴 𝗙𝗮𝗰𝘁𝘀 : 🔸 This paper is published Arxiv2022. 🔸 Official: https://lnkd.in/gNdX7AZF 🔸 Paper : https://lnkd.in/g_wqtNkQ 🤔 What is Epipolar Transformer? ☑️ "Epipolar transformer", which enables the 2D detector to leverage 3D-aware features to improve 2D pose estimation.  ------------------------------------------------------------------- 𝗜𝗠𝗣𝗢𝗥𝗧𝗔𝗡𝗖𝗘 ☑️ Learning-based Multi-View Stereo (MVS) methods warp source images into the reference camera frustum to form 3D volumes, which are fused as a cost volume to be regularized by subsequent networks.  ☑️ The fusing step plays a vital role in bridging 2D semantics and 3D spatial associations.  ☑️ However, previous methods utilize extra networks to learn 2D information as fusing cues, underusing 3D spatial correlations and bringing additional computation costs.  ☑️ Therefore, we present MVSTER, which leverages the proposed epipolar Transformer to learn both 2D semantics and 3D spatial associations efficiently.  ☑️ Specifically, the epipolar Transformer utilizes a detachable monocular depth estimator to enhance 2D semantics and uses cross-attention to construct data-dependent 3D associations along epipolar line.  ☑️ Additionally, MVSTER is built in a cascade structure, where entropy-regularized optimal transport is leveraged to propagate finer depth estimations in each stage.  ☑️ Extensive experiments show MVSTER achieves state-of-the-art reconstruction performance with significantly higher efficiency: Compared with MVSNet and CasMVSNet, our MVSTER achieves 34% and 14% relative improvements on the DTU benchmark, with 80% and 51% relative reductions in running time. MVSTER also ranks first on Tanks&Temples-Advanced among all published works.  #computervision #artificialintelligence #imageprocessing