-
-
Notifications
You must be signed in to change notification settings - Fork 5.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
YOLO Segment sigmoid()
fix
#13939
YOLO Segment sigmoid()
fix
#13939
Conversation
…chieve better segmentation edges. (#13912) Co-authored-by: Laughing <[email protected]> Co-authored-by: Glenn Jocher <[email protected]>
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #13939 +/- ##
=======================================
Coverage 70.25% 70.25%
=======================================
Files 124 124
Lines 15891 15891
=======================================
Hits 11164 11164
Misses 4727 4727
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
Signed-off-by: Glenn Jocher <[email protected]>
Signed-off-by: Glenn Jocher <[email protected]>
ultralytics 8.2.42
Segment Sigmoid() fix
ultralytics 8.2.42
Segment Sigmoid() fixultralytics 8.2.42
YOLO Segment sigmoid()
fix
Main branch results:
|
PR results
|
Signed-off-by: Glenn Jocher <[email protected]>
Signed-off-by: Glenn Jocher <[email protected]>
Signed-off-by: Glenn Jocher <[email protected]>
ultralytics 8.2.42
YOLO Segment sigmoid()
fixsigmoid()
fix
@Laughing-q this fix PR seems to make a lot of sense, and the edges are much smoother (see #13912), but segment mAP drops very slightly after implementing it. What do you think? |
@glenn-jocher I also tested yolov8s-seg and yolov8m-seg, both of them shown slightly mAP drop as well. # main
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.405
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.633
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.432
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.216
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.456
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.582
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.327
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.521
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.556
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.351
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.621
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.741
# segment-fix
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.404
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.634
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.431
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.213
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.456
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.581
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.326
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.520
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.555
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.347
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.620
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.740 yolov8s-seg: # main
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.366
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.580
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.387
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.168
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.410
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.537
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.306
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.488
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.521
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.301
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.585
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.715
# segment-fix
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.365
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.580
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.385
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.168
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.409
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.537
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.306
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.487
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.520
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.300
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.583
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.714 But the update does show better visualization, meanwhile eliminating |
@glenn-jocher Here are some visualized samples on main branch and this branch, and it looks like on the latter in some level it's fixing the "cut" issue we are having when BTW I think we are having too many |
@Laughing-q yes I think on a mathematical level this PR does make sense, it's taking all the values that range from -inf to +inf and placing the threshold at 0, so there will be a better dynamic range on the op than if we sigmoid from 0 to 1 and then do the threshold at 0.5. The mAP drops are in the Ultralytics mAP but the COCO mAP is mostly the same, and the visual results do look better, less discretization, so let's merge this. |
@zxDeepDiver PR merged! Thank you for your contributions. |
@Laughing-q yes the multiple mask processing functions are too many, I agree, I think there are at least two because of the order of processing the upsample, either before or after the mask processing and depending on if it's |
@glenn-jocher Maybe this fix can be made optional? My previously trained neural network, focused on working with ultra-small objects, shows a lower quality... |
pip3 install -U ultralytics==8.2.42 predict_t2_brain_512 predict_t2_ischemia pip3 install -U ultralytics==8.2.43 predict_t2_brain_512 predict_t2_ischemia All F1 values worse on the new version |
Hi @Alarmod, Thank you for sharing your detailed results. It appears that the recent update has impacted the performance metrics of your model, particularly the F1 scores. We understand how crucial these metrics are for your application, especially when dealing with ultra-small objects. To address this, we can consider making the recent fix optional. This would allow users to choose whether to apply the sigmoid operation based on their specific use case. In the meantime, you can revert to the previous version (8.2.42) to maintain your model's performance. If you could provide a minimum reproducible example of your code, it would help us investigate the issue further and ensure that any changes we make will not negatively impact your use case. You can find more details on how to create a reproducible example here. Additionally, please ensure that you are using the latest versions of all dependencies to rule out any compatibility issues. Thank you for your patience and understanding. We are committed to improving the performance and flexibility of our models and appreciate your feedback. |
* Display Val images per class (ultralytics#12645) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Docs: Update HUB Teams page (ultralytics#13215) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Fix ARM64 Docker building (ultralytics#13217) Co-authored-by: Glenn Jocher <[email protected]> * Fix Classifier last layer indexing (ultralytics#13219) * Update Exporter for `tf_keras` install (ultralytics#13231) * Install `cmake` with `onnxsim` (ultralytics#13222) * Add multiple lines graph support in `analytics` 8.2.26 (ultralytics#13214) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Add Prettier for YAML formatting to Ultralytics Actions (ultralytics#13236) Co-authored-by: UltralyticsAssistant <[email protected]> * Docs Solutions to Navigation Bar (ultralytics#13249) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Muhammad Rizwan Munawar <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Richard Abrich <[email protected]> * Add Data Collection and Annotation Docs Page and Preprocessing Annotated Data Docs Page (ultralytics#13253) * Update `tensorstore` link for ARM64 Docker (ultralytics#13264) Co-authored-by: Glenn Jocher <[email protected]> * Add YOLOv10 model description in docs home page (ultralytics#13265) Co-authored-by: Glenn Jocher <[email protected]> * Fix `xyxyxyxy2xywhr` for Numpy inputs (ultralytics#13273) * `ultralytics 8.2.27` replace `onnxsim` with `onnxslim` (ultralytics#12989) Co-authored-by: inisis <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: inisis <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Use new `ultralytics-thop` package (ultralytics#13282) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Revert `ultralytics-thop` to optional (ultralytics#13290) * Update HUB Integrations page in Docs (ultralytics#13292) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Docs spelling and grammar fixes (ultralytics#13307) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: RainRat <[email protected]> * Update HUB SDK Docs (ultralytics#13309) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * import `torchvision` during Predict warmup (ultralytics#13313) * `ultralytics 8.2.28` improved NMS speeds (ultralytics#13315) Signed-off-by: Glenn Jocher <[email protected]> * Fix for ValueError (expected type INT8) (ultralytics#13341) * Suppress YouTube Test `ConnectionError` (ultralytics#13357) * Add ops.py comment for `nm` = number of masks (ultralytics#13353) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Fix Docs export table from `onnxsim` to `onnxslim` (ultralytics#13324) Co-authored-by: Glenn Jocher <[email protected]> * Add area chart in `analytics` (ultralytics#13391) Co-authored-by: UltralyticsAssistant <[email protected]> * Clean up Docs pages (ultralytics#13370) Co-authored-by: Glenn Jocher <[email protected]> * Search for model metadata with TensorFlow GraphDef (ultralytics#13389) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Edge TPU export table fix (ultralytics#13420) * Add https://youtu.be/pWYiene9lYw & https://youtu.be/Unt4Lfid7aY to docs (ultralytics#13380) Co-authored-by: Glenn Jocher <[email protected]> * Update Minimum Reproducible Example (MRE) Docs page (ultralytics#13443) Signed-off-by: Glenn Jocher <[email protected]> * Update Python file headers (ultralytics#13445) Co-authored-by: Glenn Jocher <[email protected]> * Fix Paddle 2.6.0 Dockerfile install bug (ultralytics#13447) Signed-off-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.29` new fractional AutoBatch feature (ultralytics#13446) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Burhan <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Refactor Python code (ultralytics#13448) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Code Refactor for Speed and Readability (ultralytics#13450) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.30` automated tags and release notes (ultralytics#13164) Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Fix release publish action (ultralytics#13462) * Update publish.yml to handle existing tag (ultralytics#13463) * Update publish.yml (ultralytics#13464) * Docs Prettier reformat (ultralytics#13483) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Update CI to `pull_request_target` (ultralytics#13495) * Add NVIDIA Jetpack4 and Jetpack5 Docker Images (ultralytics#13100) Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Lakshantha <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.31` NVIDIA Jetpack4 and Jetpack5 Dockerfile Images (ultralytics#13496) Co-authored-by: UltralyticsAssistant <[email protected]> * Fix `jp` to `jetpack` (ultralytics#13499) * Add https://youtu.be/ydGdibB5Mds to docs (ultralytics#13555) * `ultralytics 8.2.32` Apple MPS device Autobatch handling (ultralytics#13568) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.33` pin `numpy<2.0.0` for compatibility (ultralytics#13661) Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.34` bump `ultralytics-thop>=2.0.0` (ultralytics#13662) Co-authored-by: UltralyticsAssistant <[email protected]> * Code Refactor `ruff check --fix --extend-select I` (ultralytics#13672) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Clean up unused `Silence` module (ultralytics#13674) * `ultralytics 8.2.35` add YOLOv9t/s/m models (ultralytics#13504) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Update YOLOv9 YAMLs (ultralytics#13756) Signed-off-by: Glenn Jocher <[email protected]> * Fix TensorRT Doc Overview URL (ultralytics#13759) * YOLOv9 model docs page cleanup (ultralytics#13757) Co-authored-by: Glenn Jocher <[email protected]> * Apply new Ruff actions to Python codeblocks (ultralytics#13783) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * ROS Quickstart Guide (ultralytics#13553) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Update Ruff formatting actions (ultralytics#13787) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * https://youtu.be/biIW5Z6GYl0 to docs (ultralytics#13797) * Replace https://youtu.be/biIW5Z6GYl0 with https://www.youtube.com/embed/biIW5Z6GYl0 (ultralytics#13804) * Deprecate `Silence` module in favor of `nn.Identity` (ultralytics#13785) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Laughing-q <[email protected]> * `ultralytics 8.2.36` update Ultralytics color palette (ultralytics#13808) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Update CLA Action (ultralytics#13831) * Update OpenCV image read error message (ultralytics#13822) Co-authored-by: Glenn Jocher <[email protected]> * ROS quickstart Docs page cleanup (ultralytics#13835) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Annotator `txt_color` updates (ultralytics#13842) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Muhammad Rizwan Munawar <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.37` update `temporary_modules` and Remove YOLOv9e `Silence` module (ultralytics#13819) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.38` official YOLOv10 support (ultralytics#13113) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Laughing-q <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Laughing <[email protected]> * Code refactor https://ultralytics.com/actions (ultralytics#13844) * ROS Quickstart, fixed code formatting (ultralytics#13855) Co-authored-by: UltralyticsAssistant <[email protected]> * Replace `+=` with faster list `.append()` (ultralytics#13849) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Fix ambiguous variable names (ultralytics#13864) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Alex Pasquali <[email protected]> * Add https://youtu.be/tq3FU_QczxE to docs (ultralytics#13867) Co-authored-by: Glenn Jocher <[email protected]> * Update yolov8-p6.yaml with model parameters and GFLOPs (ultralytics#13862) * Fix HUB link https://ultralytics.com/hub (ultralytics#13884) Co-authored-by: Mughees Ahmad <[email protected]> * `ultralytics 8.2.39` update `onnxslim>=0.1.31` (ultralytics#13883) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Kayzwer <[email protected]> * `ultralytics 8.2.40` refactor HUB code into callbacks (ultralytics#13896) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Adjust `numpy<2.0.0` compatibility (ultralytics#13906) Signed-off-by: Glenn Jocher <[email protected]> * Fix HUB session with DDP training (ultralytics#13103) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Burhan <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Compress ROS Guide Images (ultralytics#13914) * `ultralytics 8.2.41` fix HUB unzip subdirectory bug (ultralytics#13913) * Add YOLOv8 OpenVINO C++ Inference example (ultralytics#13839) Co-authored-by: Muhammad Amir Abdurrozaq <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Ultralytics TensorRT10 update (ultralytics#13933) Signed-off-by: Glenn Jocher <[email protected]> * Dockerfile FROM `pytorch/pytorch:2.3.1-cuda12.1-cudnn8-runtime` (ultralytics#13937) Signed-off-by: Glenn Jocher <[email protected]> * Add CLI commands for `predict` and `train` YOLOv10 models. (ultralytics#13940) Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.42` NVIDIA TensorRT 10 default (ultralytics#13943) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: DeepDiver <[email protected]> Co-authored-by: Laughing <[email protected]> * YOLO Segment `sigmoid()` fix (ultralytics#13939) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: DeepDiver <[email protected]> Co-authored-by: Laughing <[email protected]> * Add warning for inference end2end models with `augment` arg (ultralytics#13958) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Adjust box labels on right image side (ultralytics#13959) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.43` enable `classes` filter for end2end models (ultralytics#13971) Co-authored-by: Glenn Jocher <[email protected]> * Added a `max_size` parameter to the `plot_images` function (ultralytics#14002) Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Add Tips for Model Training Docs Page (ultralytics#14011) Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.44` Increase Predict dataloader robustness (ultralytics#14005) Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.45` Fix YOLOv8 `augment` inference (ultralytics#14017) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.46` fix OBB Results `xyxy` attribute (ultralytics#14020) * Add https://youtu.be/eX5ad6udQ9Q to docs (ultralytics#14077) Co-authored-by: Glenn Jocher <[email protected]> * Fix deprecation warning (ultralytics#14091) Co-authored-by: UltralyticsAssistant <[email protected]> * Add Insights on Model Evaluation and Fine-Tuning Docs Page (ultralytics#14085) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Strip `dfl_loss` from `BboxLoss` (ultralytics#14041) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.46` Results, DFL and AIGym fixes (ultralytics#14074) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: AAOMM <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Muhammad Rizwan Munawar <[email protected]> Co-authored-by: zzzer <[email protected]> Co-authored-by: Abirami Vina <[email protected]> * `ultralytics 8.2.47` YOLOv8 zero-shot action recognition example (ultralytics#13935) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.48` strip model `criterion` on save (ultralytics#14106) Signed-off-by: Glenn Jocher <[email protected]> * Default strip_optimizer() to `use_dill=False` (ultralytics#14107) Signed-off-by: Glenn Jocher <[email protected]> * Ultralytics Code Refactor https://ultralytics.com/actions (ultralytics#14109) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Update NVIDIA Jetson DeepStream Guide with YOLOv8 and Jetson Orin Support (ultralytics#14059) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> * Update FAQ.md (ultralytics#14134) Co-authored-by: UltralyticsAssistant <[email protected]> * Update Results and CFG docstrings (ultralytics#14139) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Add Docs models pages FAQs (ultralytics#14167) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Add Model Testing Guide and Best Practices for Model Deployment Guide (ultralytics#14105) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: RizwanMunawar <[email protected]> Co-authored-by: Muhammad Rizwan Munawar <[email protected]> * Add https://youtu.be/mUybgOlSxxA to docs (ultralytics#14195) Co-authored-by: Glenn Jocher <[email protected]> * Add FAQ sections to Modes and Tasks (ultralytics#14181) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Abirami Vina <[email protected]> Co-authored-by: RizwanMunawar <[email protected]> Co-authored-by: Muhammad Rizwan Munawar <[email protected]> * `ultralytics 8.2.49` fix classification `setup_model` (ultralytics#14199) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Add FAQs to Docs Datasets and Help sections (ultralytics#14211) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Fix mkdocs.yml raw image URLs (ultralytics#14213) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Burhan <[email protected]> * Fix Action Recognition Example with `torch>=2.0` (ultralytics#14232) * Add Discourse at https://community.ultralytics.com (ultralytics#14231) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.50` new Streamlit live inference Solution (ultralytics#14210) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Muhammad Rizwan Munawar <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: RizwanMunawar <[email protected]> Co-authored-by: Kayzwer <[email protected]> * Pin `tensorrt<=10.1.0` to fix `libnvinfer_builder_resource_win.so.10.2.0` error (ultralytics#14239) Signed-off-by: Glenn Jocher <[email protected]> * Dockerfile install `tensorrt-cu12==10.1.0` (ultralytics#14240) Signed-off-by: Glenn Jocher <[email protected]> * Update Pose docs with keypoint explanations (ultralytics#14248) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.51` update YOLOv9 `GITHUB_ASSETS_NAMES` (ultralytics#14261) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Allow OpenVINO export from CUDA (ultralytics#14256) Co-authored-by: Glenn Jocher <[email protected]> * Fix `end2end` attribute in `init_criterion` (ultralytics#14267) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Add Maintaining Your Computer Vision Models Docs Page (ultralytics#14304) Co-authored-by: UltralyticsAssistant <[email protected]> * `allow_empty=True` for Classify dataset class directories (ultralytics#14301) Co-authored-by: Glenn Jocher <[email protected]> * Fix Annotator PIL Image size (width, height) order (ultralytics#14227) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.52` fix CenterCrop transforms for PIL Image inputs (ultralytics#14308) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Lucas Buligon Antunes <[email protected]> * BaseTrainer with `find_unused_parameters=True` when using DistributedDataParallel() (ultralytics#14323) * Ultralytics Asset URL Update (ultralytics#14345) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.53` Heatmaps fix for empty images (ultralytics#14329) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: RizwanMunawar <[email protected]> * `ultralytics 8.2.54` resolve YouTube bug with switch to `pytubefix` (ultralytics#14354) Signed-off-by: Glenn Jocher <[email protected]> * Add https://youtu.be/fLBbyhPbWzY to docs (ultralytics#14356) * `ultralytics 8.2.55` adaptive `tflite_support` logic (ultralytics#14385) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> Co-authored-by: Nguyễn Anh Bình <[email protected]> Co-authored-by: Johnny <[email protected]> Co-authored-by: Muhammad Rizwan Munawar <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.56` Streamlit tracking app (ultralytics#14269) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> Co-authored-by: Nguyễn Anh Bình <[email protected]> Co-authored-by: Johnny <[email protected]> * Fix `TORCHVISION_0_18` for `allow_empty=True` (ultralytics#14415) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Add https://youtu.be/gX5kSRD56Gs to docs (ultralytics#14417) * Fix `model.save()` method to FP16 (ultralytics#14418) Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.57` new Solutions Tests and Docs (ultralytics#14408) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Scope `tkinter` and `streamlit` in Solutions tests (ultralytics#14426) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Fix `_process_batch()` docstrings (ultralytics#14454) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.58` FastSAM code refactor (ultralytics#14450) Co-authored-by: Glenn Jocher <[email protected]> * Fix `model` parameter in Pose, Segment dataset train examples (ultralytics#14505) * Warn on `save_hybrid=True` (ultralytics#14484) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Remove redundant assignment (ultralytics#14498) Co-authored-by: Glenn Jocher <[email protected]> * Update `FastSAM` and `SAM` docs (ultralytics#14499) Co-authored-by: Glenn Jocher <[email protected]> * Add Kaggle Integrations Docs Page (ultralytics#14487) Co-authored-by: UltralyticsAssistant <[email protected]> * Fix `Enable Tracking` Button and Optimize FPS in Streamlit Application (ultralytics#14508) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.59` use `Results.save_txt` for validation (ultralytics#14496) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Allow `test-dev2017.txt` for val (ultralytics#14519) Co-authored-by: UltralyticsAssistant <[email protected]> * Add https://youtu.be/isc06_9qnM0 to docs (ultralytics#14525) Co-authored-by: Glenn Jocher <[email protected]> * Fix NCNN multiple-volumes PNNX download bug (ultralytics#14533) * `ultralytics 8.2.60` refactor `process_mask_upsample` (ultralytics#14474) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Fix Multi-GPU trained model export (ultralytics#14551) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Fix Docs pretty `__init__.py` URLs (ultralytics#14550) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Add https://youtu.be/ziJR01lKnio to docs (ultralytics#14554) * Add Custom CLIP Model Download Path (ultralytics#14517) Co-authored-by: wangsrGit119 <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.61` fix `num_threads` for CPU training (ultralytics#14544) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Pin `pytubefix==6.3.4` for YouTube fix (ultralytics#14571) * Replace `enumerate` + index with `zip()` (ultralytics#14574) * Engine Model and Results Docs improvements (ultralytics#14564) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Felipe Parodi <[email protected]> * Add Streamlit Inference Python `model` arg (ultralytics#14563) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.62` add Explorer CLI `model` and `data` args (ultralytics#14581) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Mohammed Yasin <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Fix Docs plaintext link scan (ultralytics#14583) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> * Plaintext negative lookbehind scan (ultralytics#14601) Signed-off-by: Glenn Jocher <[email protected]> * Replace enumerate with zip in models/yolo (ultralytics#14599) * `ultralytics 8.2.63` refactor `FastSAMPredictor` (ultralytics#14582) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Laughing <[email protected]> Co-authored-by: Laughing-q <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Update format.yml for PAT (ultralytics#14608) Co-authored-by: UltralyticsAssistant <[email protected]> * Fix OBB Docs page commas error (ultralytics#14609) * Add NAS autodownload (ultralytics#14627) * Windows `torch==2.4.0` Segment `augment=True` failed test bypass (ultralytics#14637) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Fix torchvision InterpolationMode warnings (ultralytics#14632) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> * Patch `torch.load(..., weights_only=False)` to reduce warnings (ultralytics#14638) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Attempt to fix NAS models inference (ultralytics#14630) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> * Fix Streamlit Inference model suffix bug (ultralytics#14621) Co-authored-by: malopez <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Fix `torch.amp.autocast('cuda')` warnings (ultralytics#14633) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> * `ultralytics 8.2.64` YOLOv10 SavedModel, TFlite, and GraphDef export (ultralytics#14572) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * New AGLU activation module (ultralytics#14644) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Fix `torch.amp` has no attribute `GradScaler` (ultralytics#14647) Co-authored-by: UltralyticsAssistant <[email protected]> * Add warning when `cache_ram` works with classify (ultralytics#14650) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Allow `agnostic_nms` option for validation (ultralytics#14675) * `ultralytics 8.2.65` fix YouTube throttling bug (ultralytics#14684) * Remove duplicate `make_divisible` function (ultralytics#14690) Co-authored-by: UltralyticsAssistant <[email protected]> * Add https://youtu.be/_gRqR-miFPE to docs (ultralytics#14698) * Updates `save_period` to include first epoch (ultralytics#14700) Co-authored-by: Glenn Jocher <[email protected]> * Add compatible `tensorstore` versions for `aarch64` (ultralytics#14697) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.66` HUB model autodownload (ultralytics#14702) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `SETTINGS` type and key checks (ultralytics#14703) Signed-off-by: Glenn Jocher <[email protected]> * Add JetPack6 Docker for NVIDIA Jetson Orin Series (ultralytics#14707) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.67` new NVIDIA Jetson Orin Jetpack 6 Docker image (ultralytics#14740) * Simplify Dockerfile `WORKDIR` (ultralytics#14750) Signed-off-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.68` new HUB GCP region latency tests (ultralytics#14753) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Add JupyterLab Integrations Docs Page (ultralytics#14755) Co-authored-by: UltralyticsAssistant <[email protected]> * Fixed OpenVINO Docs formatting (ultralytics#14773) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Add IBM Watsonx Integrations Docs Page (ultralytics#14785) Co-authored-by: UltralyticsAssistant <[email protected]> * Update Ultralytics issue templates (ultralytics#14718) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Burhan <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Eliminate `set()` to avoid `set()` + `list()` op (ultralytics#14745) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.69` FastSAM prompt inference refactor (ultralytics#14724) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Update comet from init() to login() (ultralytics#14793) Co-authored-by: UltralyticsAssistant <[email protected]> * New Meta Segment Anything Model 2 (SAM2) Docs page (ultralytics#14794) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * SAM2 mkdocs.yml fix (ultralytics#14796) * `py-cpuinfo` Exception context manager fix (ultralytics#14814) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Add https://youtu.be/_gRqR-miFPE & https://youtu.be/CfbHwPG01cE to docs (ultralytics#14817) * `ultralytics 8.2.70` Segment Anything Model 2 (SAM 2) (ultralytics#14813) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Exported model batch size validation fix (ultralytics#14845) Co-authored-by: Glenn Jocher <[email protected]> * Fix `model` in Pose, Segment datasets train FAQ sections (ultralytics#14877) * Fixed `box_label` docstrings (ultralytics#14866) Co-authored-by: dearMOMO <[email protected]> * Add missing CLI `yolo` commands for `TASK` and `MODE` in Docs - Quickstart and CLI Guide (ultralytics#14882) Co-authored-by: Glenn Jocher <[email protected]> * Update SAM 2 docs (ultralytics#14864) * Increase Dockerfile build `max-parallel` (ultralytics#14892) * Fix incorrect CLI commands in Datasets Docs (ultralytics#14889) * Add Segment masks to YOLO-Seg labels converter (ultralytics#14557) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.71` Multinode DDP training (ultralytics#14879) Co-authored-by: Haris Rehman <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Delete Docker Cache before building Image (ultralytics#14894) Co-authored-by: Glenn Jocher <[email protected]> * ROS quickstart meta description (ultralytics#14932) * Fix the docstring of xywhr2xyxyxyxy (ultralytics#14934) Co-authored-by: Glenn Jocher <[email protected]> * Fixed `circle_label` and `text_label` docstrings (ultralytics#14909) * Fix `is_url()` and `check_disk_space()` docstrings in downloads.py (ultralytics#14923) * Update MLP module for RTDETR backward compatibility (ultralytics#14901) Co-authored-by: Glenn Jocher <[email protected]> * Corrects CONTRIBUTING.md relative link to CLA.md (ultralytics#14857) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.72` SAM 2 multiple-`bboxes` support (ultralytics#14928) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.73` Meta SAM2 Refactor (ultralytics#14867) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Ignore Vimeo 401 'unauthorized' errors (ultralytics#14980) * fix example for plotting Ray Tune history (ultralytics#14970) Co-authored-by: Glenn Jocher <[email protected]> * Update NVIDIA Jetson Docs with JetPack 6 (ultralytics#14939) Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> * Fix OpenVINO Export Docs (ultralytics#14918) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.74` add `fuse_score=True` BoT-SORT and ByteTrack arg (ultralytics#14965) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Fix Docker git permissions (ultralytics#14995) Signed-off-by: Glenn Jocher <[email protected]> * Dedicated Inference API Docs (ultralytics#14992) Co-authored-by: Glenn Jocher <[email protected]> * Update HUB Inference API Docs (ultralytics#15035) Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Sergiu Waxmann <[email protected]> * Add `allow_background_images=True` in split_dota.py (ultralytics#15037) Co-authored-by: Glenn Jocher <[email protected]> * `ultralytics 8.2.75` new Docs author profiles (ultralytics#15050) Signed-off-by: Glenn Jocher <[email protected]> * Update Docs CSS (ultralytics#15062) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Use macros for Docs tables de-duplication (ultralytics#14990) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Delete `/macros` dir from Docs site (ultralytics#15068) Co-authored-by: UltralyticsAssistant <[email protected]> * Delete macros from sitemap.xml (ultralytics#15105) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Add https://youtu.be/EeDd5P4eS6A to docs (ultralytics#15107) * Optimized SAHI video inference (ultralytics#15183) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Update `convert_segment_masks_to_yolo_seg` to support custom datasets (ultralytics#15176) * `ultralytics 8.2.76` Autobackend TensorRT/Triton Segment metadata fix (ultralytics#15185) Co-authored-by: Glenn Jocher <[email protected]> * Update Contributing guidelines (ultralytics#15373) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * Fixed multiscale preprocess_batch (ultralytics#15392) * Improve trainer DDP device handling (ultralytics#15383) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Update Conda CI (ultralytics#15443) * Update Tracker docstrings (ultralytics#15469) Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> * `ultralytics 8.2.77` new `color_mode=instance` plot arg (ultralytics#15034) Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> * Update yolov8_sahi.py (ultralytics#15470) Co-authored-by: UltralyticsAssistant <[email protected]> * Ultralytics Actions JSON, CSS and autolabel support (ultralytics#15599) Co-authored-by: Glenn Jocher <[email protected]> * Remove unnecessary assignments (ultralytics#15582) * YOLO Vision 2024 updates https://ultralytics.com/events/yolovision (ultralytics#15602) Co-authored-by: Muhammad Rizwan Munawar <[email protected]> --------- Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: Adamcode <[email protected]> Co-authored-by: UltralyticsAssistant <[email protected]> Co-authored-by: Glenn Jocher <[email protected]> Co-authored-by: Sergiu Waxmann <[email protected]> Co-authored-by: Lakshantha Dissanayake <[email protected]> Co-authored-by: Muhammad Rizwan Munawar <[email protected]> Co-authored-by: Richard Abrich <[email protected]> Co-authored-by: Abirami Vina <[email protected]> Co-authored-by: Kayzwer <[email protected]> Co-authored-by: inisis <[email protected]> Co-authored-by: inisis <[email protected]> Co-authored-by: RainRat <[email protected]> Co-authored-by: czy10383 <[email protected]> Co-authored-by: kubade-ashish <[email protected]> Co-authored-by: Burhan <[email protected]> Co-authored-by: Ultralytics Assistant <[email protected]> Co-authored-by: Paula Derrenger <[email protected]> Co-authored-by: Ivor Zhu <[email protected]> Co-authored-by: Ahmed Mahfouz <[email protected]> Co-authored-by: Lakshantha <[email protected]> Co-authored-by: alexwine36 <[email protected]> Co-authored-by: Laughing <[email protected]> Co-authored-by: Francesco Mattioli <[email protected]> Co-authored-by: Laughing-q <[email protected]> Co-authored-by: xiaoluohao <[email protected]> Co-authored-by: Alex Pasquali <[email protected]> Co-authored-by: Mughees Ahmad <[email protected]> Co-authored-by: Erlangga Yudi Pradana <[email protected]> Co-authored-by: Muhammad Amir Abdurrozaq <[email protected]> Co-authored-by: DeepDiver <[email protected]> Co-authored-by: jackwolfey <[email protected]> Co-authored-by: bobyard-com <[email protected]> Co-authored-by: zzzer <[email protected]> Co-authored-by: AAOMM <[email protected]> Co-authored-by: fatih c. akyon <[email protected]> Co-authored-by: RizwanMunawar <[email protected]> Co-authored-by: JF Chen <[email protected]> Co-authored-by: Antônio Martos Harres <[email protected]> Co-authored-by: Alejandro Casanova <[email protected]> Co-authored-by: Alexis IMBERT <[email protected]> Co-authored-by: Sheffey <[email protected]> Co-authored-by: Lucas Buligon Antunes <[email protected]> Co-authored-by: Chia-Hsiang Tsai <[email protected]> Co-authored-by: Nguyễn Anh Bình <[email protected]> Co-authored-by: Johnny <[email protected]> Co-authored-by: Haonan Liu <[email protected]> Co-authored-by: Aryan Jassal <[email protected]> Co-authored-by: zhiqiang yang <[email protected]> Co-authored-by: Ariel Kukulanski <[email protected]> Co-authored-by: Mohammed Yasin <[email protected]> Co-authored-by: suke <[email protected]> Co-authored-by: wangsrGit119 <[email protected]> Co-authored-by: Felipe Parodi <[email protected]> Co-authored-by: rulosant <[email protected]> Co-authored-by: Miguel Angel Lopez <[email protected]> Co-authored-by: malopez <[email protected]> Co-authored-by: Hassan Ghaffari <[email protected]> Co-authored-by: Kevin <[email protected]> Co-authored-by: Jan Knobloch <[email protected]> Co-authored-by: dearMOMO <[email protected]> Co-authored-by: Haris Rehman <[email protected]> Co-authored-by: Haris Rehman <[email protected]> Co-authored-by: Zeel B Patel <[email protected]> Co-authored-by: Maxi <[email protected]> Co-authored-by: Sergiu Waxmann <[email protected]> Co-authored-by: Galasnow <[email protected]> Co-authored-by: alanZee <[email protected]>
Signed-off-by: Glenn Jocher <[email protected]> Co-authored-by: DeepDiver <[email protected]> Co-authored-by: Laughing <[email protected]>
…chieve better segmentation edges. (#13912)
🛠️ PR Summary
Made with ❤️ by Ultralytics Actions
🌟 Summary
This PR simplifies mask processing logic and enhances testing in the Ultralytics repository.
📊 Key Changes
pytest-cov
for test coverage reporting.process_mask_*
functions inultralytics/utils/ops.py
to remove unnecessary sigmoid operations and adjusted the mask threshold condition.🎯 Purpose & Impact
pytest-cov
helps in measuring test coverage, leading to higher code quality and reliability.