Qing Guo, Wei Feng, Ivor Tsang, Song Wang, Xiaoguang Li, Di Lin, Rabab Abdelfattah
We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Fine-tuning Inpainting-pretrained Naive Encoder-decoders with different amounts of shadow data: 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% Naive En.-De. (a) (b) BMNet Figure 1: (a) Comparison with SOTA methods on ISTD+ dataset[24] at shadow (Y-axis) and shadow-free (X-axis) regions. We also present the results of a naive encoder-decoder network (Naive En.-De.) trained with different strategies: (1) training on inpainting data and fine-tuning on different amounts of shadow data; (2) only training on all shadow data. (b) Displays two examples of four methods and we use white arrows to highlight the main advantages.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2302.05361")
get_code_for_paper("2302.05361")
have("2302.05361")
The run record, dated, one paper per request, free:
curl https://syntology.ai/api/ran/2302.05361.json
A badge for a README (the split and the date, never a ratio):
[](https://syntology.ai/paper/2302.05361)
Connect an agent — have() is free.