Vladimir Iglovikov

Just run your code from https://github.com/kornia/kornia-rs/blob/main/py-kornia/benchmark/bench_resize.py Would be happy to check a version with an updated setup for a benchmark :)

Do you have comparison with other libraries, somthing similar to: https://github.com/albumentations-team/albucore/blob/main/benchmark/results/uint8_3.md and if resize in kornia-rs is not broken, as it is in torchvision, tensorflow or OpenCV it would be...

> To integrate this custom_dataset with YOLOv8's training function train(data=...), you would typically set up a DataLoader and then integrate it within the training process. For YOLOv8, ensure that you...

@glenn-jocher In the code that you provided, there no single line related to Ultralytics. It is just a general pytorch dataloader and training pipeline. Could you please share on how...

@glenn-jocher Here is doc about using Albumentations for Object Detection with HuggingFace https://huggingface.co/docs/transformers/main/en/tasks/object_detection I wanted to create something similar for UltraLytics, meaning using existing implemented functionality, but replacing DataLoader with...

Thanks for the answer What about defining custom Detection trainer like ```python from ultralytics.models.yolo.detect.train import DetectionTrainer class AlbumentationsTrainer(DetectionTrainer): def build_dataset(self, img_path, mode="train", batch=None): return ``` and pass it as: ```python...

Thanks! Will work on this ASAP.

Related to https://github.com/albumentations-team/albumentations/issues/1811

@GeorgePearse I could implement this, the question is how to do it in a useful way. For that I need to start training object detection models, which I am not...

@GeorgePearse do you have it as a part of the working training pipeline?