A repository that shares tuning results of trained models generated by TensorFlow / Keras. Post-training quantization (Weight Quantization, Integer Quantization, Full Integer Quantization, Float16 Quantization), Quantization-aware training. TensorFlow Lite. OpenVINO. CoreML. TensorFlow.js. TF-TRT. MediaPipe. [.tflite,.h5,.pb,saved_model,tfjs,tftrt,mlmodel,.xml/.bin]
tensorflow
dbface
quantization
super-resolution
coreml
posenet
sound-classification
tensorflow-lite
deeplabv3
tensorflowjs
openvino
edgetpu
mobilenetv3
mediapipe
blazeface
efficientdet
facemesh
objectron
blazepose
tf-trt
-
Updated
Nov 24, 2020 - Python
The counter variable which updates the bottom linear progress bar on the camera screen needs to have a different instance in every model inference.
Also, the inference of each yoga pose should stop and the user should be asked to redo a yoga pose once the user has completed one yoga pose (i.e. when the counter for a yoga pose reaches 100, the inference should be stopped and he should be asked wh