Open blake-espeland opened 2 years ago
@jared3831 is having issues with the datatypes/shapes of input tensors. The repo we're using expects a json file that lists the image file and the annotations. Either there is a conversion between these two shapes, or it may be easier to just start over from scratch.
Merged #10, however, this doesn't include @jared3831's changes.
Finally included @jared3831's changes in #12
Implementation complete, ironing things out now
Based on this repository.![image](https://user-images.githubusercontent.com/46688118/171271625-bb9b957b-f035-419d-9c83-817eac9aaa67.png)
This offers an insane improvement in terms of performance over the current anchor-based YOLOv5. This feature is mainly the implementation of a new detection head.