Add YOLOX object detection #24
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
antimora
reviewed
Mar 8, 2024
- Pre-trained weights are from COCO (README) - Remove training outputs TODO - Current example uses YOLOX-Tiny (Nano WIP)
laggui
commented
Mar 19, 2024
| } | ||
|
|
||
| /// YOLOX-X pre-trained weights. | ||
| pub enum YoloxX { |
There was a problem hiding this comment.
Style-wise I am still unsure about the enum variants due to the PascalCase... Names like YoloxS and YoloxX look weird to me, so feel free to leave suggestions if you have any!
There was a problem hiding this comment.
Ahh, that's fine to me - it seems super minor.
antimora
approved these changes
Mar 20, 2024
| } | ||
|
|
||
| /// YOLOX-X pre-trained weights. | ||
| pub enum YoloxX { |
There was a problem hiding this comment.
Ahh, that's fine to me - it seems super minor.
nathanielsimard
approved these changes
Mar 20, 2024
|
@laggui ready to merge? |
Actually I think we're waiting for the Burn release so we can use the crates version and announce some things. I'll let @nathanielsimard decide when we merge this. |
- Removed init_with methods - Fixed empty MaxPool2d vec initialization
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Closes issue #21
This PR adds a Burn implementation for the YOLOX object detection models. Implementation and weights are based on the official implementation.
TODO
NMS tensor implementationRequiresargsortWe should compare timings w/ the current naive implementation to see if it is worth it since we don't have the nms kernel to perform the computations even more efficientlyI scratched the NMS tensor implementation. Not required at this time.