Repository navigation
Add owned frame preprocessing and optional bounded video decoding - #1149
Draft
Baiju Meswani (baijumeswani) wants to merge 1 commit into
Draft
Baiju Meswani (baijumeswani) wants to merge 1 commit into
Baiju Meswani (baijumeswani) wants to merge 1 commit into
Conversation
Provide owned RGB frames, authoritative timestamps, resolved patch preprocessing and source-compatible resize modes. Add opt-in shared FFmpeg loading with bounded selection, decode work and geometry. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: a32500b6-be93-4206-a257-11406b00a3f3
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Add reusable owned RGB frame inputs and optional bounded local-file video loading to the Extensions preprocessing API. Model-specific prompt formatting, temporal positions, scheduling and feature replacement remain in GenAI.
OCOS_ENABLE_VIDEO_DECODERloading with an externally supplied shared FFmpeg 8.x SDK. Caller-frame preprocessing remains usable without the decoder.Dependencies
This draft is based on the bounded-tokenizer prerequisite #1148. Its diff contains media preprocessing and decoding only.
Validation
The committed Linux decoder-enabled candidate passes all 45 targeted processor/video-loading cases, including image regressions, temporal packing, geometry/ownership failures, VFR and B-frame selection, audio ignoring, orientation, truncation and quota handling.
Decoder-disabled execution is also validated before publication, preserving frame support and API symbols without a file-decoding dependency.
The integrated GenAI candidate has passing Qwen/Flash functional matrices and reduced-Ling lifecycle evidence. These are not claims of complete source numerical qualification or full Ling execution.
Draft status
Windows shared-library ABI, codec SDK packaging and runtime execution remain to be performed on the separate Windows machine. This draft does not claim complete model quality qualification.