mirrors/gstreamer
mirrors/gstreamer
1
1
Fork 0
mirror of https://gitlab.freedesktop.org/gstreamer/gstreamer.git synced 2024-06-09 01:29:28 +00:00
Code Issues 1 Projects Releases Wiki Activity

doc: remove details about sinkpad caps and batch

Browse source
This commit is contained in:
Daniel Morin 2024-05-14 21:50:22 -04:00
parent 8f3b7e4fd9
commit aec7a85128
1 changed files with 1 additions and 14 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
subprojects/gst-docs/markdown/additional/design/machine-learning-analytics.md
View file

@ -268,20 +268,7 @@ make negotiation of tensor-decoder in particular difficult.
### Inference Sinkpad(s) Capabilities
Sinkpad capability, before been constrained based on model, can be any
media type, including `tensor` . Note that multiple sinkpads can be present.
#### Batch Inference
To support batch inference `tensor` media type need to be used. Batching is
a method used to spread the fixed time cost of scheduling work on some
accelerator (GPU). Multiple samples (buffers) are aggregated into a batch that
is pushed to the inference. When batching is used, output tensor will also contain
analytics results in the form of a batch. Un-batching require information on how
the batch was formed: buffer timestamp, buffer source, media type, buffer caps.
A batch can be formed from a single source, forming the batch with time multiplexing
of from multiple sources, time and source multiplexing. Once multiplexed, the
batch can be pushed to the inference element as a tensor media.
TODO: Describe TensorBatchMeta
media type.
### Inference Srcpad(s) Capabilities