I'm trying to use a custom YOLOv1 model for person detection (by modifying the example). However,despite the model being relatively small, I'm getting excessive RAM usage and overflowing ie:
Failed to resize buffer. Requested: 399776, available 376504, missing: 23272
I have attached my model, both in tflite format and its converted format (using neutron converter). Would someone be able to point out the cause of itsexcessive RAM use and suggest possible optimizations?
