search

sekilab / VehicleOrientationDataset

The vehicle orientation dataset is a large-scale dataset containing more than one million annotations for vehicle detection with simultaneous orientation classification using a standard object detection network.
MIT License
51 stars 6 forks source link

Yolo Tiny #1

Open jaydubal opened 1 year ago

jaydubal commented 1 year ago

Hello, is it possible to have yolo3 tiny version of this? Thanks!

ashutoshIITK commented 1 year ago

@jaydubal Do you need a pre-trained model for YOLOv3-tiny? Can you send me the project link?

jaydubal commented 1 year ago

Hello Ashutosh, thanks for your reply! I don't have any project here on github. I am a C# freelance developer and I intend to use YOLOv3-tiny with a CPU (only). I am using https://github.com/AlturosDestinations/Alturos.Yolo#pre-trained-dataset for the object detector. I wish to capture front or rear information. Your provided model files work but they will need a GPU, which I don't have. Thanks.

could you please suggest why following is not being detected (using yolo 4) Car1

while a cropped version of it does work.

Car2

So, is it expecting image with some specific aspect ratio?