Hi everyone, I want to share my github repo deploying image classification onto the embedded Jetson Nano platform, improving performance by optimizations from frozen graph model tensorflow1, kernel fusion, and FP32/FP16 precision. we will guide you inference and realtime with CPU, GPU, FP32 and FP16 then make some compare between different inference and Also guide you how to fine tuning and deploy model to tensorrt engine on Jetson nano or Google Colab
github : GitHub - SokPhanith/tf1_tensorrt_image_classification_jetson_nano: Image classification TensorRT
example inference with image
example inference by fine-tuning with flowers dataset and deploy
example inference by fine-tuning with custom dataset vgg_19 on google colab and deploy
