Hey all,
Im using DeepStream to detect objects from video taken by a drone.
I wish to train a ‘yolov4 tiny’ network, using TAO so it could detect these specific objects that are seen by the drone.
Is it possible only recorded a video using Deepstream (and save the stream as videos on the SD card) without any image analytics, so after that, I could break the videos into frames and train the network? If yes, could you please provide an example?
There is no update from you for a period, assuming this is not an issue anymore.
Hence we are closing this topic. If need further support, please open a new one.
Thanks