Wang Ao

> Same issue in my model, can we freely set a smaller threshold for YOLOv10 to detect more small object? @pseacrest Yes, we think so.

@NickHerrig Thanks! Would you mind checking if the following results in your local environment match that of the demo? Thank you! > > * Loss of accuracy resulting from conversion...

@SkalskiP Thank you very much! The results of the demo seem to be still different from our local environment. We are investigating this. We will get back to you once...

@SkalskiP @NickHerrig We found that the inference results seem to be not the same in our codebase and Roboflow Inference with the same onnx file. Here is a minimal example...

@NickHerrig Thanks a lot for your great efforts! Is the different preprocessing step between `inference` and `yolo cli` expected?

@SkalskiP It seems that we and @NickHerrig have identified the root cause. One of the reasons is that roboflow inference invokes the NMS in the postprocessing of YOLOv10, which is...

@SkalskiP The PR https://github.com/roboflow/inference/pull/437 has been merged. The results of roboflow Inference and our local environment are the same now. Would you mind updating the `inference` version in the `requirements.txt`...

@SkalskiP Friendly ping :) Thanks!

@SkalskiP We opened a PR in https://huggingface.co/spaces/SkalskiP/YOLO-ARENA/discussions/2 to update the `inference` version of the [HF Space](https://huggingface.co/spaces/SkalskiP/YOLO-ARENA)? Would you mind taking a look? Thanks a lot!

Thanks very much for your contribution! But we encounter error when running the demo locally. ``` ValueError: Cannot process this value as an Image, it is of type: ``` Could...