Image rescaling could be trivial but we want to make sure rescaling will not cause a bottleneck that slows down the ML model's performance or affects accuracy. There are multiple libraries that can perform image rescaling on the Jetson Nano such as OpenCV, jetson-utils, and probably TensorFlow. Is there one approach that achieves the right balance of performance and complexity?
Image rescaling could be trivial but we want to make sure rescaling will not cause a bottleneck that slows down the ML model's performance or affects accuracy. There are multiple libraries that can perform image rescaling on the Jetson Nano such as OpenCV, jetson-utils, and probably TensorFlow. Is there one approach that achieves the right balance of performance and complexity?