Raja-89 c60ac8f256 avfilter/dnn: implement batching for Torch backend ...

Add batch processing support to the LibTorch DNN backend, following
the same pattern used by the OpenVINO backend.

Key changes:
- Add batch_size AVOption (range 1-1000, default 1) to DnnContext
- Accumulate lltasks in the queue, trigger inference when batch_size
  lltasks are ready (matching the OpenVINO batch execution pattern)
- Pre-allocate a single contiguous memory buffer for the entire batch
  in fill_model_input_th() to avoid unnecessary tensor copies
- Split batched output in infer_completion_callback() and dispatch
  each slice to its corresponding task

Tested with:
ffmpeg -f lavfi -i testsrc=duration=5:size=640x480:rate=25 -vf format=rgb24,dnn_processing=dnn_backend=torch:model=dummy_model.pt:batch_size=4 -f null -

Signed-off-by: Raja Rathour <imraja729@gmail.com>

2026-07-18 02:56:18 +00:00

..

dnn_backend_common.c

…

dnn_backend_common.h

…

dnn_backend_onnx.c

avfilter/dnn: add ONNX Runtime backend with GPU execution provider support

2026-06-22 20:06:40 +08:00

dnn_backend_openvino.c

…

dnn_backend_tf.c

avfilter/dnn_backend_tf: fix ctx async field access

2026-03-20 02:22:06 +00:00

dnn_backend_torch.cpp

avfilter/dnn: implement batching for Torch backend

2026-07-18 02:56:18 +00:00

dnn_interface.c

avfilter/dnn: implement batching for Torch backend

2026-07-18 02:56:18 +00:00

dnn_io_proc.c

…

dnn_io_proc.h

…

Makefile

avfilter/dnn: add ONNX Runtime backend with GPU execution provider support

2026-06-22 20:06:40 +08:00

queue.c

…

queue.h

…

safe_queue.c

…

safe_queue.h

…