Thanks to visit codestin.com
Credit goes to github.com

Skip to content

optimized - H.265 tuned submission (score: 3.32)#22

Closed
ArmanJR wants to merge 1 commit into
commaai:masterfrom
ArmanJR:submission/optimized
Closed

optimized - H.265 tuned submission (score: 3.32)#22
ArmanJR wants to merge 1 commit into
commaai:masterfrom
ArmanJR:submission/optimized

Conversation

@ArmanJR

@ArmanJR ArmanJR commented Apr 4, 2026

Copy link
Copy Markdown
Contributor

submission name:

optimized

upload zipped archive.zip

https://github.com/ArmanJR/comma_video_compression_challenge/releases/download/submission-optimized-v1/archive.zip

report.txt

=== Evaluation config ===
  batch_size: 16
  device: mps
  num_threads: 2
  prefetch_queue_depth: 4
  report: submissions/optimized/report.txt
  seed: 1234
  submission_dir: submissions/optimized
  uncompressed_dir: videos
  video_names_file: public_test_video_names.txt
=== Evaluation results over 600 samples ===
  Average PoseNet Distortion: 0.31108347
  Average SegNet Distortion: 0.00480904
  Submission file size: 1,619,264 bytes
  Original uncompressed size: 37,545,489 bytes
  Compression Rate: 0.04312806
  Final score: 100*segnet_dist + √(10*posenet_dist) + 25*rate = 3.32

does your submission require gpu for evaluation (inflation)?

no

did you include the compression script? and want it to be merged?

no

additional comments

I'm still thinking and experimenting on this challenge but wanted to submit my score so far. It's not the best, but hopefully I can come up with better solutions in the future.

Approach

H.265 (libx265) with slow preset, CRF 25, inter-frame coding (bframes=8, ref=6), and parameters tuned for the scoring models. Decode uses yuv420_to_rgb() from frame_utils.py with bilinear upscale to match the evaluation pipeline.

@github-actions

github-actions Bot commented Apr 4, 2026

Copy link
Copy Markdown

Thanks for the submission @ArmanJR! 🤏

A maintainer will review your PR shortly.

To run the evaluation, a maintainer will trigger the eval workflow with your PR number.

@github-actions github-actions Bot requested a review from YassineYousfi April 4, 2026 20:23
@ArmanJR

ArmanJR commented Apr 4, 2026

Copy link
Copy Markdown
Contributor Author

@ylevental I literally acknowledged that in my submission. But thanks for taking the time to not read it and still comment. Truly groundbreaking analysis.

@ylevental

ylevental commented Apr 4, 2026

Copy link
Copy Markdown

@ylevental I literally acknowledged that in my submission. But thanks for taking the time to not read it and still comment. Truly groundbreaking analysis.

I see now in additional comments. Apologies for that. I guess another issue with AI is that can make people overly combative

@ArmanJR

ArmanJR commented Apr 4, 2026

Copy link
Copy Markdown
Contributor Author

@ylevental I literally acknowledged that in my submission. But thanks for taking the time to not read it and still comment. Truly groundbreaking analysis.

I see now in additional comments. Apologies for that. I guess another issue with AI is that can make people overly combative

No worries. Funny how AI keeps getting the blame for a comment you typed.

@YassineYousfi

YassineYousfi commented Apr 4, 2026

Copy link
Copy Markdown
Collaborator

@ArmanJR please rename to an informative name instead of just "optimized"
something like h265_tuned

@github-actions

github-actions Bot commented Apr 4, 2026

Copy link
Copy Markdown

Eval Results: optimized

=== Evaluation config ===
  batch_size: 16
  device: cpu
  num_threads: 2
  prefetch_queue_depth: 4
  report: submissions/optimized/report.txt
  seed: 1234
  submission_dir: submissions/optimized
  uncompressed_dir: /home/runner/work/comma_video_compression_challenge/comma_video_compression_challenge/videos
  video_names_file: /home/runner/work/comma_video_compression_challenge/comma_video_compression_challenge/public_test_video_names.txt
=== Evaluation results over 600 samples ===
  Average PoseNet Distortion: 0.31108448
  Average SegNet Distortion: 0.00480906
  Submission file size: 1,619,264 bytes
  Original uncompressed size: 37,545,489 bytes
  Compression Rate: 0.04312806
  Final score: 100*segnet_dist + √(10*posenet_dist) + 25*rate = 3.32

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@YassineYousfi YassineYousfi Awaiting requested review from YassineYousfi

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@ArmanJR @ylevental @YassineYousfi