This is an automated email from the git hooks/post-receive script. Git pushed a commit to branch master in repository ffmpeg.
commit 1e0279143db99d7324b17f9784b3229122269b38 Author: Raja-89 <[email protected]> AuthorDate: Tue Aug 4 23:47:01 2026 +0530 Commit: guoyejun <[email protected]> CommitDate: Sun Aug 9 02:13:24 2026 +0000 doc/filters: document async, nireq, and batch_size options for dnn_processing Add documentation for three previously undocumented common DNN options: - async: toggle async inference (default enabled) - nireq: number of concurrent inference requests - batch_size: frames per inference request Also add usage examples for the Libtorch backend showing basic CPU inference and batch inference pipelines. Signed-off-by: Raja Rathour <[email protected]> --- doc/filters.texi | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) diff --git a/doc/filters.texi b/doc/filters.texi index 4cedd53b26..51885fddde 100644 --- a/doc/filters.texi +++ b/doc/filters.texi @@ -12278,6 +12278,26 @@ Roll back to sync execution if the backend does not support async. For tensorflow backend, you can set its configs with @option{sess_config} options, please use tools/python/tf_sess_config.py to get the configs of TensorFlow backend for your system. +@item async +Use DNN async inference (default: 1). When enabled, frames are submitted +to the backend without blocking the caller. Supported by the Torch, +OpenVINO, and TensorFlow backends; the ONNX Runtime backend ignores +this option. + +@item nireq +Number of concurrent inference requests (default: 0, meaning auto). +When set to 0, the backend automatically determines an appropriate +number of requests. Higher values increase throughput by allowing +multiple frames to be in flight simultaneously. + +@item batch_size +Number of frames to batch into a single inference request (default: 1). +Values greater than 1 cause the backend to accumulate frames and +concatenate them along the batch dimension before running the model. +This can significantly increase throughput on GPUs at the cost of +slightly higher per-frame latency. Currently supported by the OpenVINO +and Torch backends. + @item device Set the device to run the model. For the ONNX Runtime backend this selects the execution provider: @code{cpu} (default), @code{cuda} (NVIDIA GPU), @@ -12329,6 +12349,19 @@ throughout the pipeline: -y output.mp4 @end example +@item +Process rgb24 frames with a TorchScript model using the Libtorch backend: +@example +./ffmpeg -i input.jpg -vf format=rgb24,dnn_processing=dnn_backend=torch:model=my_model.pt output.jpg +@end example + +@item +Process rgb24 frames with the Libtorch backend using batch inference +(four frames per request): +@example +./ffmpeg -i input.mp4 -vf format=rgb24,dnn_processing=dnn_backend=torch:model=my_model.pt:batch_size=4 -y output.mp4 +@end example + @end itemize @section drawbox _______________________________________________ ffmpeg-cvslog mailing list -- [email protected] To unsubscribe send an email to [email protected]
