Songyou Peng, Marc Pollefeys, Jiaqi Chen, Weining Ren, Boyang Sun, Zihan Zhu
We lifted 2 functions out of this paper's own repositories and ran 1 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| cvg/nerf-on-the-go | — | 1 of 2 |
| Function | Status | Where it lives |
|---|---|---|
| _robustnerf_inner_patch_mask | Ran | cvg/nerf-on-the-go/internal/robustnerf.py code served (permissive licence) · get_code("6964d8f3b1fcfaee") |
| robustnerf_mask | Not yet run | cvg/nerf-on-the-go/internal/robustnerf.py code served (permissive licence) · get_code("353e4db9afd42470") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Figure 1. NeRF On-the-go. Given casually captured image sequences or videos in the wild as inputs, the goal of this paper is to train a NeRF for static scenes and effectively remove all dynamic elements in the scenes (cars, trams, pedestrians, etc), i.e. distractors. Unlike existing methods such as NeRF-W [23] and RobustNeRF [35], which produce imperfect results, our method leverages the predicted uncertainty maps to effectively remove those distractors. This results in high-fidelity novel view synthesis on challenging dynamic scenes.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2405.18715")
get_code_for_paper("2405.18715")
have("2405.18715")
Connect an agent — have() is free.