livepeer / livepeer/research

GPU-based Transcoding Verification

Open
#12 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
No language data
Stars
7
Forks
1
PR merge metrics
No merged PRs in 30d

Description

GPUs are 2-4x more efficient than CPUs at video transcoding. However, they use floating point calculation, and therefore don't fit into the current verification scheme. Instead of verifying the transcoding result based on a hash that requires bit-wise exact output, we can use a combination of video fingerprints, gop size, resolution, and video quality to verify a transcode output.

This also unlocks multi-threaded transcoding on CPUs.

Areas of research include:
* [Perceptual Hashing](https://en.wikipedia.org/wiki/Perceptual_hashing), video fingerprinting.
* Video quality measurement (should try [VMAF](https://github.com/Netflix/vmaf))
* Allowing multiple transcoding settings (and verification methods)
* Allowing broadcaster to pick verification method

Follow the discord discussion [here](http://discord.gg/Z4DXx63).

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reviewing the issue's proposed research areas: perceptual hashing, video fingerprinting, VMAF, GOP size, resolution, and video quality measurement. The issue names no repository files, tests, or entry points; done would require a defined verification design that supports GPU and multithreaded CPU transcoding.

Written by the indexing model from the issue text.

Assessment

Domain
audio-video-rtc
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.