Thanks for your answer.The logic we put in place supposes that the maximum input size for a tiny yolov2 is 608, else the model will be too big to fit in the target Ram and flash.So try to not use a resolution bigger than 608.By the way in the next re...
Hello,Can you share your terminal output.Can you confirm that you have a nvidia GPU properly configured to use tensorflow_gpu on your machine?Best Regards.
Hello,It seems that you modified the yaml file structure which caused an issue.This is an example of correctly structured yaml file that you can use: https://github.com/STMicroelectronics/stm32ai-modelzoo/blob/main/object_detection/datasets/dataset_c...
Hello,Please note that the model you are using is indeed quantized, but it has float32 output.In the C code, you need to make sure to handle the output data as float32.Best Regards.