-
-
Notifications
You must be signed in to change notification settings - Fork 6.5k
Insights: ultralytics/ultralytics
Overview
Could not load contribution data
Please try again later
2 Releases published by 1 person
-
v8.3.48 v8.3.48 - `ultralytics 8.3.48` split PyPI publish jobs for security (#18111)
published
Dec 8, 2024 -
v8.3.49 v8.3.49 - `ultralytics 8.3.49` Docker images `uv pip install` (#18115)
published
Dec 10, 2024
17 Pull requests merged by 9 people
-
Add https://youtu.be/qE-dfbB5Sis to docs
#18207 merged
Dec 12, 2024 -
Add YOLO11 audio podcast
#18174 merged
Dec 11, 2024 -
Fix SAM CUDA hard-code
#18153 merged
Dec 11, 2024 -
Fix untrained warning when training from yaml
#18168 merged
Dec 11, 2024 -
Fix validation callbacks not triggered during OBB model training
#18175 merged
Dec 11, 2024 -
Fix deleted author profile
#18180 merged
Dec 11, 2024 -
Removed duplicate IMX500 docs reference
#18178 merged
Dec 11, 2024 -
ultralytics 8.3.49
Docker imagesuv pip install
#18115 merged
Dec 10, 2024 -
Git pull docs before updating
#18163 merged
Dec 10, 2024 -
Remove Google Drive test
#18162 merged
Dec 10, 2024 -
Change index to start from 1 by default in
predictions.json
#18140 merged
Dec 10, 2024 -
Update
torchvision
compatibility table#18117 merged
Dec 9, 2024 -
Update YOLOv5 export table links
#18130 merged
Dec 9, 2024 -
Update Jetson Doc with DLA info
#18128 merged
Dec 9, 2024 -
Bump astral-sh/setup-uv from 3 to 4 in /.github/workflows
#18123 merged
Dec 9, 2024 -
ultralytics 8.3.48
split PyPI publish jobs for security#18111 merged
Dec 8, 2024 -
OpenVINO broken link fix
#18107 merged
Dec 8, 2024
12 Pull requests opened by 8 people
-
Use `uv` in `check_requirements`
#18116 opened
Dec 9, 2024 -
Prevent BatchNorm stats from being changed on model initialization from `yaml`
#18149 opened
Dec 10, 2024 -
Improve the robustness of `resample_segments`
#18171 opened
Dec 11, 2024 -
Add argument validation for export
#18185 opened
Dec 11, 2024 -
Add support for loading and inferencing on aws-neuron and aws-neuronx
#18199 opened
Dec 12, 2024 -
Dwell time
#18205 opened
Dec 12, 2024 -
Fix `model.save()` error when loading from yaml
#18212 opened
Dec 13, 2024 -
Add callback "on_val_postprocess_end" to validator
#18214 opened
Dec 13, 2024 -
Add type for `train` arguments
#18221 opened
Dec 13, 2024 -
Add https://youtu.be/j0MOGKBqx7E to docs
#18222 opened
Dec 13, 2024 -
Add redirects for yolov5 docs pages
#18226 opened
Dec 13, 2024
43 Issues closed by 37 people
-
Training on Pose and bounding boxes and validating only on bounding boxes
#17917 closed
Dec 15, 2024 -
How to extract mAP50 mAP50-95 in each class from model.val() for Yolov11
#18209 closed
Dec 13, 2024 -
Segmentation of overlapping objects
#17965 closed
Dec 13, 2024 -
Exporting trained YOLOv6 model crashes
#18225 closed
Dec 13, 2024 -
How to Merge or use two weights at the same time
#16200 closed
Dec 13, 2024 -
Why latest version is downloading `yolo11n.pt` while training `yolov8*.pt` (or any other model)?
#18220 closed
Dec 13, 2024 -
Question about Image Processing after Converting to CoreML
#18203 closed
Dec 13, 2024 -
How to stop ultralytics from setting default running folder
#18193 closed
Dec 13, 2024 -
Fine-Tuning using other dataset
#18069 closed
Dec 13, 2024 -
Yolov11n conversion to TensorFlow Lite INT8 detect no object in any image
#18210 closed
Dec 12, 2024 -
Yolov8 Docker Image >v8.2.27 not saving Results anymore
#16425 closed
Dec 12, 2024 -
YOLOv11 architecture for classification
#18182 closed
Dec 12, 2024 -
Using 'mps' - view size is not compatible with input tensor's size and stride. Use .reshape(...) instead
#18184 closed
Dec 12, 2024 -
Bug```
#18187 closed
Dec 11, 2024 -
YOLOv11 to ONNX (with NMS and dynamic batch)
#18172 closed
Dec 11, 2024 -
Region counter
#16292 closed
Dec 11, 2024 -
Cite YOLOv11
#18148 closed
Dec 11, 2024 -
4xT4 (16GB), YOLOV11L-seg, Error :- torch OOM Error
#17854 closed
Dec 11, 2024 -
YOLOv10 NMS not working on RK3588
#18129 closed
Dec 11, 2024 -
YOLO World fine tuning trains on wrong classes as compared to what is provided
#18142 closed
Dec 11, 2024 -
Detecting an specific group of people on an image
#18100 closed
Dec 10, 2024 -
When running segment/val.py file, an incorrect predictions.json file is generated
#18136 closed
Dec 10, 2024 -
How do you combine yolov8 with tracking algorithms other than botsort and bytetrack?
#14477 closed
Dec 10, 2024 -
Are these metrics normal? Visdrone YOLO11n
#18031 closed
Dec 10, 2024 -
How get annotations from yolov8seg predictions
#18141 closed
Dec 10, 2024 -
yolo segmentation mask from binary mask.
#17174 closed
Dec 10, 2024 -
Two sizes of images are used for training. Training speed reduced by 8 times
#16494 closed
Dec 10, 2024 -
ERR when transform YOLOv11 to tflite file
#18135 closed
Dec 10, 2024 -
Convert keras to tflte
#18134 closed
Dec 9, 2024 -
yolov11 export to tflite does not work on Android GPU
#17837 closed
Dec 9, 2024 -
FileNotFoundError: [WinError 3] 系统找不到指定的路径。: '/tmp/ultralytics_runner'
#18132 closed
Dec 9, 2024 -
engine model predict total time about 250ms
#18071 closed
Dec 9, 2024 -
CLI and SDK dont have the same imgsz order
#17293 closed
Dec 9, 2024 -
How to load model with load_state_dict ?
#18097 closed
Dec 9, 2024 -
ModuleNotFoundError: No module named 'ultralytics.nn.C2f_EMA'
#18121 closed
Dec 9, 2024 -
weird training process&result during training yolo11 with 8.3.28
#18058 closed
Dec 9, 2024 -
Found discrepency between actual dataset size and ultralytics docs value - Open Images v7
#18082 closed
Dec 9, 2024 -
Discrepancy between what's in GitHub and what's been published to PyPI for v8.3.41
#18027 closed
Dec 8, 2024
66 Issues opened by 63 people
-
Is it possible to use both heads for keypoint detection and instance segmentation?
#18241 opened
Dec 15, 2024 -
Breakpoint training cannot update results properly.
#18239 opened
Dec 15, 2024 -
is there a tool can split custom obb data to train and val,evenly divided by classes
#18238 opened
Dec 15, 2024 -
why does test dataset take no effect ?
#18236 opened
Dec 14, 2024 -
The YOLO11S model cannot infer large images, while YOLOV8S can
#18235 opened
Dec 14, 2024 -
Ask how yolov11 achieves segmentation.
#18234 opened
Dec 14, 2024 -
i want to chanage pratmter in traing while resuming
#18233 opened
Dec 14, 2024 -
DOTA dataset
#18232 opened
Dec 14, 2024 -
Save class-wise metrics table as a .csv along within the runs folder, after running `.train` and `.val`
#18230 opened
Dec 13, 2024 -
Inconsistent `ch` attribute not working as expected
#18229 opened
Dec 13, 2024 -
Is it possible to change the order of detection from top to bottom?
#18227 opened
Dec 13, 2024 -
Installing `ultralytics` with poetry causes numpy version complaint
#18224 opened
Dec 13, 2024 -
if onnx inference is on cpu?
#18223 opened
Dec 13, 2024 -
Hyperparameters in custom hyp.yaml Not Applied During Model Validation
#18219 opened
Dec 13, 2024 -
Export ONNX with V11 produce bad result
#18218 opened
Dec 13, 2024 -
YOLOv8-OBB How to obtain the detection accuracy of large and small targets
#18217 opened
Dec 13, 2024 -
YOLO11L-Seg pretrain large dataset
#18213 opened
Dec 13, 2024 -
SAM2 fails with long videos
#18211 opened
Dec 12, 2024 -
Error in Fake Filesystem Simulation During YOLOv8 Training
#18208 opened
Dec 12, 2024 -
Saving yolov6n crashes
#18206 opened
Dec 12, 2024 -
Is there a way to have SAM2VideoPredictor do a sequence of images and frame-by-frame tracking?
#18204 opened
Dec 12, 2024 -
Val statistics for different box sizes in detect mode for custom dataset
#18202 opened
Dec 12, 2024 -
Filtering predictions with conf_thres in non_max_suppression
#18201 opened
Dec 12, 2024 -
Why the model uses more vram in the larger size images?
#18200 opened
Dec 12, 2024 -
How can I troubleshoot when my code stops running after displaying "engine/trainer"?
#18198 opened
Dec 12, 2024 -
the code for training yoloworld can really reproduce the results? have you test it?
#18196 opened
Dec 12, 2024 -
triton-inference-server low throughput
#18191 opened
Dec 12, 2024 -
Image window stuck using model.predict show=True
#18190 opened
Dec 12, 2024 -
yolov8-pose
#18189 opened
Dec 12, 2024 -
Wrong boxes location in train and val labels
#18186 opened
Dec 11, 2024 -
How to display the real segmented images labeled on the COCO dataset
#18177 opened
Dec 11, 2024 -
Save predicted images to same directory, instead of incemented runs directory.
#18176 opened
Dec 11, 2024 -
How to Look at Feature Importance
#18170 opened
Dec 11, 2024 -
How to Identifying Misclassified Images During YOLOv8 Model Validation
#18169 opened
Dec 11, 2024 -
How to convert coco panoptic segmentation dataset to yolo format
#18167 opened
Dec 11, 2024 -
SAMURAI implementation
#18165 opened
Dec 10, 2024 -
TensorRT & YOLO 11
#18161 opened
Dec 10, 2024 -
Metrics + Loss calculations
#18160 opened
Dec 10, 2024 -
confusion matrix
#18159 opened
Dec 10, 2024 -
YOLOv11-Seg : How to make input data from torch.Tensor or numpy.array, not from yaml
#18158 opened
Dec 10, 2024 -
workspace parameter in export method does not accept None despite documentation
#18157 opened
Dec 10, 2024 -
implementation of LDConv
#18156 opened
Dec 10, 2024 -
Yolov8 Pose Landmark Detection, not getting full success
#18155 opened
Dec 10, 2024 -
Change number of epochs after when resuming training.
#18154 opened
Dec 10, 2024 -
how to get hyperparameter when model.tune completed
#18152 opened
Dec 10, 2024 -
reproduce the performance of YOLOV8s,YOLOV9s,YOLOV10s on COCO.
#18150 opened
Dec 10, 2024 -
Not selecting class weights manually
#18147 opened
Dec 10, 2024 -
Can I use YOLO's pose detection to detect keypoints in charts?
#18145 opened
Dec 10, 2024 -
YOLOv8 with SAHI
#18144 opened
Dec 10, 2024 -
How can I replace backbone with swin transformer
#18137 opened
Dec 9, 2024 -
Hello, may I ask why the output size changes during the training and evaluation phases?
#18133 opened
Dec 9, 2024 -
Yolo label segment resampling breaks labels with holes in them
#18131 opened
Dec 9, 2024 -
help/
#18127 opened
Dec 9, 2024 -
How to Assign the Same ID to same group
#18125 opened
Dec 9, 2024 -
size_problem
#18124 opened
Dec 9, 2024 -
PYQT5+YOLO
#18120 opened
Dec 9, 2024 -
OBB probiou replace
#18119 opened
Dec 9, 2024 -
'DetectionModel' object has no attribute 'reg_max'
#18118 opened
Dec 9, 2024 -
Issue with Installation on Python 3.13
#18114 opened
Dec 8, 2024 -
Yolov8 and Conv.py
#18110 opened
Dec 8, 2024 -
The inference time of the int8 exported ONNX model is abnormal.
#18109 opened
Dec 8, 2024
220 Unresolved conversations
Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.
-
Try uv pip install for RPi Benchmarks
#17912 commented on
Dec 10, 2024 • 2 new comments -
extremely low metrics for yolov8-obb training with DOTA1.5 dataset
#16232 commented on
Dec 10, 2024 • 0 new comments -
In the segmentation task, the self-made data labels have accurate true value contours, but the contours are not accurate during training
#16178 commented on
Dec 11, 2024 • 0 new comments -
Train custom Dataset with fixed H:480 but different Width: 600 - 850
#16166 commented on
Dec 11, 2024 • 0 new comments -
Fast computation methods in yolov8 tracking
#16164 commented on
Dec 11, 2024 • 0 new comments -
How to use masks from segmentation model as SAM2 prompt
#17943 commented on
Dec 11, 2024 • 0 new comments -
export yolo 11 in ONXX with options doesn't change the accuracy and confidence score. are these option working in yolov11?
#16549 commented on
Dec 11, 2024 • 0 new comments -
Training iteratively with dynamic resumes
#18015 commented on
Dec 11, 2024 • 0 new comments -
Losses are nan upon resuming training with resume=True [Yolov8]
#17282 commented on
Dec 11, 2024 • 0 new comments -
Make Enough FPS with Multiple Models for Multiple Stream
#16160 commented on
Dec 12, 2024 • 0 new comments -
Proper CoreML Conversion for YOLOv10
#16157 commented on
Dec 12, 2024 • 0 new comments -
update read.me file to detail VisualStudio imports for /examples/YOLOv8-ONNXRuntime-CPP project
#16145 commented on
Dec 12, 2024 • 0 new comments -
Parameters in C2f and issues with logging output
#16142 commented on
Dec 12, 2024 • 0 new comments -
The x y width height of the yolov10 model output looks like a pixel value. What is the size of the model output?
#16136 commented on
Dec 12, 2024 • 0 new comments -
Questions about object detection
#16140 commented on
Dec 12, 2024 • 0 new comments -
CUDA tutorial
#16119 commented on
Dec 12, 2024 • 0 new comments -
crash when use ObjectCounter in background terminal environment without a graphical user interface
#16099 commented on
Dec 12, 2024 • 0 new comments -
How to set a fixed learning rate for YOLOV8?
#16098 commented on
Dec 12, 2024 • 0 new comments -
YOLOX vs YOLOV10
#16096 commented on
Dec 12, 2024 • 0 new comments -
how to set val batchsize=1 when training with yolov8 to detect?
#16095 commented on
Dec 12, 2024 • 0 new comments -
Anyway to boost yolo performance on Jetson Orin?
#17640 commented on
Dec 12, 2024 • 0 new comments -
Extend device support.
#16930 commented on
Dec 12, 2024 • 0 new comments -
ICYMI - YOLO Vision 2024 VOD is on YouTube
#16786 commented on
Dec 12, 2024 • 0 new comments -
Object Detection Model Produces Poor Results for Thin and Long Objects
#9768 commented on
Dec 12, 2024 • 0 new comments -
Effective Techniques for Quantizing YOLO Models (v8, v11) to Achieve Size Compression Under 1MB
#17535 commented on
Dec 12, 2024 • 0 new comments -
KeyError when Running custom YOLOv8 Segmentation Model on Edge TPU with Raspberry Pi 5
#16094 commented on
Dec 13, 2024 • 0 new comments -
RuntimeError: shape '[1, 67, -1]' is invalid for input of size 38400
#16088 commented on
Dec 13, 2024 • 0 new comments -
Integrating YOLOv8n model for Real-Time Webcam Detection and Image Upload/Camera Capture in a Flutter Web Application
#16083 commented on
Dec 13, 2024 • 0 new comments -
Validation Losses are immediately increasing (detection)
#16081 commented on
Dec 13, 2024 • 0 new comments -
Warning generated by RT-DETR training
#16231 commented on
Dec 10, 2024 • 0 new comments -
metrics中的fitness函数
#16229 commented on
Dec 10, 2024 • 0 new comments -
Pruned model showing same params
#16227 commented on
Dec 10, 2024 • 0 new comments -
Custom Callbacks for ClearML Failing During Multi-GPU Training
#18070 commented on
Dec 10, 2024 • 0 new comments -
some bug with RandomPerspective
#15953 commented on
Dec 10, 2024 • 0 new comments -
TaskAlignedAssigner doesn't take into account `reg_max` which affects prediction of large objects
#11634 commented on
Dec 10, 2024 • 0 new comments -
How to implement a multi task detector for parallel detection and classification
#17824 commented on
Dec 10, 2024 • 0 new comments -
export YOLO11 onnx with NMS
#16732 commented on
Dec 10, 2024 • 0 new comments -
Domain adaptation for YOLOv8
#18060 commented on
Dec 10, 2024 • 0 new comments -
Core ML nms=True' is not available for end2end models. Forcing 'nms=False'.
#16927 commented on
Dec 10, 2024 • 0 new comments -
YOLOv5n training output display error
#17893 commented on
Dec 10, 2024 • 0 new comments -
Metrics calculation
#16226 commented on
Dec 11, 2024 • 0 new comments -
The conf composition during the inference process in YOLOV8
#16222 commented on
Dec 11, 2024 • 0 new comments -
How to use Tensorrt 10 in C++API to action a segment task
#16221 commented on
Dec 11, 2024 • 0 new comments -
High class imbalance
#16217 commented on
Dec 11, 2024 • 0 new comments -
YOLOv8 OpenVINO C++ inference
#16212 commented on
Dec 11, 2024 • 0 new comments -
Performance Discrepancy in YOLOv8n Inference Speed Using TensorRT on laptop RTX 4060
#16211 commented on
Dec 11, 2024 • 0 new comments -
RuntimeError: The size of tensor a (8400) must match the size of tensor b (16800) at non-singleton dimension 2
#16210 commented on
Dec 11, 2024 • 0 new comments -
A question about how to make the input of layer zero as the input of other layers
#16209 commented on
Dec 11, 2024 • 0 new comments -
yolo-pose training results highly inconsistent with the results from the validation scripts
#16208 commented on
Dec 11, 2024 • 0 new comments -
YOLOV8 Instance segmentation results
#16203 commented on
Dec 11, 2024 • 0 new comments -
deafult.yaml Issues
#16202 commented on
Dec 11, 2024 • 0 new comments -
Filtering on classes
#16198 commented on
Dec 11, 2024 • 0 new comments -
Can I train pose-estimation without indicating bounding box?
#16193 commented on
Dec 11, 2024 • 0 new comments -
AttributeError: 'dict' object has no attribute 'box' when trying to train YOLOv8 on pytorch lightning using v8DetectionLoss.
#16189 commented on
Dec 11, 2024 • 0 new comments -
labeling
#16181 commented on
Dec 11, 2024 • 0 new comments -
Deploying the YOLOv10 model on Android using the TFLite library gives inaccurate results?
#16179 commented on
Dec 11, 2024 • 0 new comments -
[Experiment] Adds Soft-NMS
#15993 commented on
Dec 8, 2024 • 0 new comments -
Multi label classification Support
#16015 commented on
Dec 14, 2024 • 0 new comments -
New Base Formatter for better user experience
#16037 commented on
Dec 8, 2024 • 0 new comments -
Removed unwanted local code
#16247 commented on
Dec 8, 2024 • 0 new comments -
Adds model weights pruning
#16256 commented on
Dec 8, 2024 • 0 new comments -
Adds new CLI command for model info
#16436 commented on
Dec 12, 2024 • 0 new comments -
Paddle to `[export]` optional dependencies
#16526 commented on
Dec 8, 2024 • 0 new comments -
Updated stale actions rules
#16563 commented on
Dec 8, 2024 • 0 new comments -
Combinatorial (All-pairs) testing
#16570 commented on
Dec 8, 2024 • 0 new comments -
Better dataset error handling
#16634 commented on
Dec 8, 2024 • 0 new comments -
Removes try-except from `torch_utils/profile`
#16750 commented on
Dec 13, 2024 • 0 new comments -
Improve the performance of the OBB model on square objects.
#16851 commented on
Dec 14, 2024 • 0 new comments -
Test Smart inference Mode
#17340 commented on
Dec 8, 2024 • 0 new comments -
[Experiment] Add support for the Pytorch OpenCL backend
#17396 commented on
Dec 9, 2024 • 0 new comments -
[Example] YOLOv8-Classification-ONNXRuntime-Python
#17421 commented on
Dec 9, 2024 • 0 new comments -
[EXP] Adds `uv run pytest` for tests
#17663 commented on
Dec 13, 2024 • 0 new comments -
Remove CSV from supported inference sources
#17703 commented on
Dec 8, 2024 • 0 new comments -
Improve DDP CUDA memory child processes cleanup
#17719 commented on
Dec 15, 2024 • 0 new comments -
Add optional groupname to W&B
#17781 commented on
Dec 9, 2024 • 0 new comments -
Added Checkbox to Question Issue template
#17803 commented on
Dec 8, 2024 • 0 new comments -
Yolo multilabel
#17809 commented on
Dec 9, 2024 • 0 new comments -
Fixed epoch number metadata
#17852 commented on
Dec 8, 2024 • 0 new comments -
fix: Parse onnx file channel error
#17911 commented on
Dec 10, 2024 • 0 new comments -
Eliminate code repetition by using a function.
#18078 commented on
Dec 15, 2024 • 0 new comments -
Update base.py
#18102 commented on
Dec 10, 2024 • 0 new comments -
Expand solutions FAQs
#18106 commented on
Dec 9, 2024 • 0 new comments -
Avoid incrementing counter for unconfirmed tracks and add `min_hits` parameter
#18108 commented on
Dec 12, 2024 • 0 new comments -
Training YOLO on thermal Images (Gray scale).
#16079 commented on
Dec 13, 2024 • 0 new comments -
plus the new categories base on the last.pt
#18033 commented on
Dec 13, 2024 • 0 new comments -
The model size and the amount of parameters decreased after pruning, but the inference time did not change significantly
#5238 commented on
Dec 13, 2024 • 0 new comments -
Fails to convert model to int8. runs out of memory everytime
#17406 commented on
Dec 13, 2024 • 0 new comments -
Knowledge distillation with yolo11
#17013 commented on
Dec 13, 2024 • 0 new comments -
TensorRT Export Method Restricts Workspace Memory to 4GB
#16162 commented on
Dec 14, 2024 • 0 new comments -
Exporting yolov10n.pt int8 does not appear to fully quantize, so when compiling for edgetpu, very few operations are mapped to TPU.
#16074 commented on
Dec 14, 2024 • 0 new comments -
About DDP training
#16069 commented on
Dec 14, 2024 • 0 new comments -
YOLOv8 for RGBD data
#3432 commented on
Dec 14, 2024 • 0 new comments -
Missing GPU support in the ultralytics/ultralytics:latest-jetson-jetpack5 image
#17345 commented on
Dec 14, 2024 • 0 new comments -
simple trainning of obb object doesn't work
#18099 commented on
Dec 14, 2024 • 0 new comments -
A higher MAP50, yet cannot predict classes for some test data. Is 100 epochs an overkill in such a case?
#17666 commented on
Dec 14, 2024 • 0 new comments -
Add new task: Relationship Detection :rocket:
#16411 commented on
Dec 15, 2024 • 0 new comments -
Integrating a GAN into YOLOv8 Architecture
#16067 commented on
Dec 15, 2024 • 0 new comments -
last_features
#16872 commented on
Dec 15, 2024 • 0 new comments -
The optimizer's best learning rate and momentum
#17460 commented on
Dec 15, 2024 • 0 new comments -
[Example] YOLOv8-NCNN-Android sample project using yolov8s-seg model.
#5251 commented on
Dec 8, 2024 • 0 new comments -
Support mask augmentation with albumentations
#6624 commented on
Dec 12, 2024 • 0 new comments -
YOLOv8-World autobackground class
#8258 commented on
Dec 8, 2024 • 0 new comments -
feat(Detection): add pos_weight parameter in BCEWithLogitsLoss
#8620 commented on
Dec 8, 2024 • 0 new comments -
add mIoU metric for segmentation model (fix CLA signing issue)
#9109 commented on
Dec 13, 2024 • 0 new comments -
Replace `thop` with `torch.profiler` for FLOPs
#12634 commented on
Dec 8, 2024 • 0 new comments -
New: Ultralytics YOLO-Human
#12702 commented on
Dec 8, 2024 • 0 new comments -
Add support of Ascend NPU devices
#13269 commented on
Dec 10, 2024 • 0 new comments -
[CONCEPT] Introduces ModelMeta class
#14706 commented on
Dec 8, 2024 • 0 new comments -
[Experiment] NMS with class probas
#15966 commented on
Dec 8, 2024 • 0 new comments -
Sorted args alphabetically
#15989 commented on
Dec 8, 2024 • 0 new comments -
Ultralytics runner keeps running in background after python execution finished
#18037 commented on
Dec 8, 2024 • 0 new comments -
Modifying training and evaluation for ROI-specific detection
#16440 commented on
Dec 9, 2024 • 0 new comments -
onnx 模型推理结果和pt推理结果不一致?
#16430 commented on
Dec 9, 2024 • 0 new comments -
Can we just train new data without training old data when new data is available ? with the result to have the model predicts both the old and the new one ?
#16434 commented on
Dec 9, 2024 • 0 new comments -
YOLOv8-Pose Custom Training
#16429 commented on
Dec 9, 2024 • 0 new comments -
How to print accuracy, mean Average Precision (mAP), Intersection over Union (IoU), Urgent problem please
#16424 commented on
Dec 9, 2024 • 0 new comments -
yolo and coco
#16419 commented on
Dec 9, 2024 • 0 new comments -
How to obtain the class confidence of each bbox
#16418 commented on
Dec 9, 2024 • 0 new comments -
The confusion matrix generated by model validation is inconsistent with the validation summary results
#16412 commented on
Dec 9, 2024 • 0 new comments -
How to save the printed log output to a file
#16408 commented on
Dec 9, 2024 • 0 new comments -
When I train yolov8l-cls, Closing dataloader mosaic is not work.
#16407 commented on
Dec 9, 2024 • 0 new comments -
add epochs training
#16402 commented on
Dec 9, 2024 • 0 new comments -
How to Resize window when webcam streaming
#16397 commented on
Dec 9, 2024 • 0 new comments -
Add all class softmax probability scores to results object
#16396 commented on
Dec 9, 2024 • 0 new comments -
corrupt JPEG data: premature end od data segment | training yolov8m
#16389 commented on
Dec 9, 2024 • 0 new comments -
training width and height support non-square?
#16388 commented on
Dec 9, 2024 • 0 new comments -
I want to get embeddings of each detected object
#16385 commented on
Dec 9, 2024 • 0 new comments -
How to use the Visdrone-VID dataset in the RT-DETR model?
#16380 commented on
Dec 9, 2024 • 0 new comments -
Transferred 162/419 items from pretrained weights
#16369 commented on
Dec 9, 2024 • 0 new comments -
Change evaluation period in training
#16368 commented on
Dec 9, 2024 • 0 new comments -
export tflite,error No module named 'keras.src.engine
#16365 commented on
Dec 9, 2024 • 0 new comments -
Is there any way to port the versions of YOLO v7 or v8 to darknet?
#16361 commented on
Dec 9, 2024 • 0 new comments -
A question about the ConfusionMatrix
#16360 commented on
Dec 9, 2024 • 0 new comments -
Deos yolov8 segmenation training early stopping considers val/seg_losss?
#16358 commented on
Dec 9, 2024 • 0 new comments -
The keypoint predictions are not connected.
#16357 commented on
Dec 9, 2024 • 0 new comments -
Pose mAP computation
#16351 commented on
Dec 9, 2024 • 0 new comments -
Fine Tuning a PreTrained YOLOv8 model
#16356 commented on
Dec 9, 2024 • 0 new comments -
Train across multiple machines
#16350 commented on
Dec 9, 2024 • 0 new comments -
How can I train yolo to label 1 class per image without bounding box?
#16514 commented on
Dec 9, 2024 • 0 new comments -
benckmark of VisDrone
#16513 commented on
Dec 9, 2024 • 0 new comments -
Fish Species Classification and Dataset
#16512 commented on
Dec 9, 2024 • 0 new comments -
Why can't Yolo predict until the edge of the picture?
#16510 commented on
Dec 9, 2024 • 0 new comments -
Error when training a YOLO detection model
#16509 commented on
Dec 9, 2024 • 0 new comments -
Must the number of categories be the same across training data when training for world detection? And I don't understand why the max sample num is set to min(nc,80).
#16508 commented on
Dec 9, 2024 • 0 new comments -
Add the ability to give multiple tags for an image for classification
#16507 commented on
Dec 9, 2024 • 0 new comments -
Details about yolov10 raw output without post processing
#16502 commented on
Dec 9, 2024 • 0 new comments -
yolo
#16499 commented on
Dec 9, 2024 • 0 new comments -
Do all the bits of the image used for learning have to be the same?
#16498 commented on
Dec 9, 2024 • 0 new comments -
GPU Usage keeps fluctuating during inference
#16497 commented on
Dec 9, 2024 • 0 new comments -
Combining script.py and best.pt files into .exe to use as a detection program, using pyinstaller to merge the files. What did I do wrong that I get an error when opening the files?
#16496 commented on
Dec 9, 2024 • 0 new comments -
@Wang-Yu-Qing hello there! 👋
#16490 commented on
Dec 9, 2024 • 0 new comments -
No result.png is generated after training, and it cannot be validated with the trained best.pt
#16484 commented on
Dec 9, 2024 • 0 new comments -
Any tool for annotation of the OBB bounding box
#16479 commented on
Dec 9, 2024 • 0 new comments -
YOLOv8-obb map
#16480 commented on
Dec 9, 2024 • 0 new comments -
Error while training on a dataset that I have trained earlier
#16471 commented on
Dec 9, 2024 • 0 new comments -
When will Yolov10 Pose be available?
#16467 commented on
Dec 9, 2024 • 0 new comments -
batch_size should be a positive integer value, but got batch_size=0.0
#16465 commented on
Dec 9, 2024 • 0 new comments -
train rt-detr ended up with half of the mAP50 of the others
#16463 commented on
Dec 9, 2024 • 0 new comments -
Nvidia DALI preprocessing
#16460 commented on
Dec 9, 2024 • 0 new comments -
Core ML exports, python 3.12 support
#16459 commented on
Dec 9, 2024 • 0 new comments -
onnx2tf export, python 3.12 support
#16458 commented on
Dec 9, 2024 • 0 new comments -
Instance segmentation result at pixel level
#16455 commented on
Dec 9, 2024 • 0 new comments -
The pre-training model is trained
#16454 commented on
Dec 9, 2024 • 0 new comments -
Support RT-DETR v2
#16444 commented on
Dec 9, 2024 • 0 new comments -
AttributeError: Can't get attribute 'NoneType' on <module 'builtins' (built-in)>
#16443 commented on
Dec 9, 2024 • 0 new comments -
Pose Estimation for DeepFashion2
#16349 commented on
Dec 9, 2024 • 0 new comments -
RGB infrared multimodal input
#16474 commented on
Dec 10, 2024 • 0 new comments -
Can't use ultralytics with bazel because of the jaraco.text dependency
#16473 commented on
Dec 10, 2024 • 0 new comments -
Why is the FP16 model 's computation time shorter than the INT8 model?
#16461 commented on
Dec 10, 2024 • 0 new comments -
TFLite Micro support
#16422 commented on
Dec 10, 2024 • 0 new comments -
DDP: multi node training erro
#16406 commented on
Dec 10, 2024 • 0 new comments -
how can I export onnx when I have dynamic block
#16391 commented on
Dec 10, 2024 • 0 new comments -
关于训练图片包含很多检测目标
#16381 commented on
Dec 10, 2024 • 0 new comments -
在yolov8-obb中导出onnx每一个conv都会后面跟一个sigmod在训练的时候怎么删除,这个删除的地方在哪里
#16379 commented on
Dec 10, 2024 • 0 new comments -
Question about the difference between yaml of yolov9 in the original author and yamy of yolov9 in Ultralytics
#16342 commented on
Dec 10, 2024 • 0 new comments -
[BUG] ultralytics[explorer] install fails on Python 3.12
#16331 commented on
Dec 10, 2024 • 0 new comments -
输出标签少于图片数量
#16299 commented on
Dec 10, 2024 • 0 new comments -
Retraining from best.pt or last.pt?
#16298 commented on
Dec 10, 2024 • 0 new comments -
Custom Data and RKNN
#16284 commented on
Dec 10, 2024 • 0 new comments -
Train Yolov8m Incremental
#16283 commented on
Dec 10, 2024 • 0 new comments -
How to obtain the number of samples for each class in the training set
#16281 commented on
Dec 10, 2024 • 0 new comments -
yolov8 default.yaml Hyperparameters
#16279 commented on
Dec 10, 2024 • 0 new comments -
How to parse the object detection results from deepsparse pipeline
#16277 commented on
Dec 10, 2024 • 0 new comments -
Error in Multiple GPU Training AWS
#16275 commented on
Dec 10, 2024 • 0 new comments -
Is it possible to stratify metrics?
#16268 commented on
Dec 10, 2024 • 0 new comments -
The meaning of the two training maps
#16266 commented on
Dec 10, 2024 • 0 new comments -
Ground truth masks that contain holes in training of segmentation tasks
#16263 commented on
Dec 10, 2024 • 0 new comments -
How to use bfloat for inference with a pytorch pt weight?
#16261 commented on
Dec 10, 2024 • 0 new comments -
How can I find `pred_dist` function definition?
#16258 commented on
Dec 10, 2024 • 0 new comments -
yolov8_pose
#16248 commented on
Dec 10, 2024 • 0 new comments -
In a YOLOv8 segmentation task, does resizing images affect the accuracy of the ground truth mask, and can it cause the ground truth polygon to shift?
#16240 commented on
Dec 10, 2024 • 0 new comments -
Custom YOLOv9e-seg model underperforming YOLOv8x-seg?
#16236 commented on
Dec 10, 2024 • 0 new comments -
set cuda device not work for predicting
#16233 commented on
Dec 10, 2024 • 0 new comments -
Is there any way to get the output of my customized encoder?
#16348 commented on
Dec 9, 2024 • 0 new comments -
Which one is better model trained on single object images or multiple object images?
#16340 commented on
Dec 9, 2024 • 0 new comments -
How can i know 'similarity of detection' and 'similatiy of detection roi'
#16345 commented on
Dec 9, 2024 • 0 new comments -
Why image size is changing when model.predict?
#16339 commented on
Dec 9, 2024 • 0 new comments -
YOLOV8x classification model top_1 accuracy results are giving true accuracy?
#16332 commented on
Dec 9, 2024 • 0 new comments -
How to training with EMA in model.train()?
#16329 commented on
Dec 9, 2024 • 0 new comments -
How does the process of YOLOv8 predict the class
#16325 commented on
Dec 9, 2024 • 0 new comments -
How to optimize Yolov8 to detect smaller objects
#16324 commented on
Dec 9, 2024 • 0 new comments -
change dataset for model
#16323 commented on
Dec 9, 2024 • 0 new comments -
Absolute rotation with OBB
#16322 commented on
Dec 9, 2024 • 0 new comments -
yolov8 has removed the Object Score branch, but how does it distinguish foreground from background?
#16318 commented on
Dec 9, 2024 • 0 new comments -
Tuning with Raytune doesn't save checkpoints nor plots
#16317 commented on
Dec 9, 2024 • 0 new comments -
Code Issue
#16315 commented on
Dec 9, 2024 • 0 new comments -
Set model parameters occurs no change
#16307 commented on
Dec 9, 2024 • 0 new comments -
Custom YOLOv8 Architecture - Fine Tuning vs. Training from scratch
#16310 commented on
Dec 9, 2024 • 0 new comments -
Can we add null images (images without any label) in training and validation set for training any model?
#16305 commented on
Dec 9, 2024 • 0 new comments -
So is yolov11 nms-free or not?
#18065 commented on
Dec 9, 2024 • 0 new comments -
model perform well but model track on video perform bad
#18104 commented on
Dec 9, 2024 • 0 new comments -
Triton Inference Server Deploy Problem
#18092 commented on
Dec 9, 2024 • 0 new comments -
"num_inputs <= kMaxInputNum was not true." running YOLOv11n on tflite_micro_runtime
#18083 commented on
Dec 9, 2024 • 0 new comments -
YOLOv8 Segmentation End2End using ONNXRuntime
#16029 commented on
Dec 9, 2024 • 0 new comments -
Example code non-functional with YOLO11 models
#17181 commented on
Dec 9, 2024 • 0 new comments -
Add Support for Metrics Handling During DDP Training
#18080 commented on
Dec 9, 2024 • 0 new comments -
Metrics doubt
#16201 commented on
Dec 9, 2024 • 0 new comments -
Difference in Desktop Tflite inference and Android TFlite inference
#16487 commented on
Dec 10, 2024 • 0 new comments -
Use ONNX Runtime to implement the deployment of YOLO-v8 Pose on Android Studio
#16481 commented on
Dec 10, 2024 • 0 new comments -
Adding Custom modules How do I configure yaml files correctly
#16476 commented on
Dec 10, 2024 • 0 new comments