Megvii-BaseDetection / Megvii-BaseDetection/BEVDepth
Why not detach depth when multiplying features with discrete depth
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 881
- Forks
- 124
- PR merge metrics
- No merged PRs in 30d
Description
https://github.com/Megvii-BaseDetection/BEVDepth/blob/main/bevdepth/layers/backbones/base_lss_fpn.py#L527
Hello, Why not using detach() to depth when multiplying discrete depth with features, if depth is detached, only depth loss will be backward to depth net, and detection loss will not be backward to depth net, will the depth estimation be more acurrate?
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start at bevdepth/layers/backbones/base_lss_fpn.py#L527 and trace how discrete depth is multiplied with features during training. Then inspect the depth and detection loss paths to determine whether detaching depth changes gradient flow; compare the behavior with and without detachment in the existing training setup. Done means documenting an evidence-based answer or a maintainer-approved change.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- computer-vision, machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100