What Synthesis Is Missing: Depth Adaptation

Integrated With Weak Supervision for Indoor Scene Parsing

Objective

Targeting semantic segmentation, our goal is to enable full supervision performance without human-labelled pixel-level annotations. Our approach exploits synthetic data where feasible and integrates weak supervision where necessary.

Method

Material

Paper (preprint) Paper (final) Supplementary Poster

Supplementary provides: Algorithm Listing, Depth Restoration details, Visual Results, Network Quantization Results

Contact

Keng-Chi Liu*, Yi-Ting Shen, Jan P. Klopp, Liang-Gee Chen

Graduate Institute of Electrical Engineering, National Taiwan University

Cite

If you find our work helpful for your research, please consider citing it:

@inproceedings{liu19whatsynthesisismissing,
        author = {Keng-Chi Liu and Yi-Ting Shen and Jan P. Klopp and Liang-Gee Chen},
        booktitle = {Proceedings of the International Conference on Computer Vision ({ICCV})},
        title = {What Synthesis Is Missing: Depth Adaption Integrated With Weak Supervision for Indoor Scene Parsing},
        year = {2019}}