Tianxing Feng

Tianxing Feng

Auto-generated profile, based on publicly available information

Tianxing Feng

Peking University,School of Electronic and Computer Engineering

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
10 October 2019 in Journal of Materials Science
Verified
Journal Article
Closed access
PDF file
YTYingjie Tang
Yingjie Tang
HZHao Zhou
Hao Zhou
XSXiupeng Sun
Xiupeng Sun
TFTianxing Feng
Tianxing Feng
XZXinya Zhao
Xinya Zhao
Yanchao MaoYanchao Mao
Yanchao Mao

1 October 2023
Verified
Conference Paper
Closed access
KXKaiqiang Xiong
Kaiqiang Xiong
RPRui Peng
Rui Peng
ZZZhe Zhang
Zhe Zhang
TFTianxing Feng
Tianxing Feng
JJJianbo Jiao
Jianbo Jiao
RWRonggang Wang
Ronggang Wang
Unsupervised Multi-View Stereo (MVS) methods have achieved promising progress recently. However, previous methods primarily depend on the photometric consistency assumption, which may suffer from two limitations: indistinguishable regions and view-dependent effects, e.g., low-textured areas and reflections. To address these issues, in this paper, we propose a new dual-level contrastive learning approach, named CL-MVSNet. Specifically, our model integrates two contrastive branches into an unsupervised MVS framework to construct additional supervisory signals. On the one hand, we present an image-level contrastive branch to guide the model to acquire more context awareness, thus leading to more complete depth estimation in indistinguishable regions. On the other hand, we exploit a scene-level contrastive branch to boost the representation ability, improving robustness to view-dependent effects. Moreover, to recover more accurate 3D geometry, we introduce an ℒ0.5 photometric consistency loss, which encourages the model to focus more on accurate points while mitigating the gradient penalty of undesirable ones. Extensive experiments on DTU and Tanks&Temples benchmarks demonstrate that our approach achieves state-of-the-art performance among all end-to-end unsupervised MVS frameworks and outperforms its supervised counterpart by a considerable margin without fine-tuning.