This app uses:
PyTorch Vision
PyTorch Model: FasterRCNN_ResNet50_FPN
PS: click submit after image uploaded

The model is adjusted to only detect car and traffic light. Example output: