ageitgey / ageitgey/face_recognition

Face Recognition CNN only uses 1940 MB of GPU memory on RTX 2080 Ti

Open
#862 4 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
56.8k
Forks
13.7k
PR merge metrics
No merged PRs in 30d

Description

  • face_recognition version: 1.2.3
  • Python version:3.6.8
  • Operating System: Ubuntu 18.04

HI, i have an Logitech c920 on a resolution of 1080P, and i have an RTX 2080 Ti with 11 GB of GPU memory, but i run the cnn version of face_recognition on cv2.VideoCapture in 1920x1080 (Webcam Full Resolution) and is a quite slow, the process only uses 1940 mb of 11000 mb. I am trying to use 1080p resolution and push all the power of the GPU

I Already compile dlib with CUDA, and i already try one example of multiprocessing and nothing change.

Contributor guide

Open the contributing guide

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 reported face_recognition CNN usage with cv2.VideoCapture at 1920x1080 and the attempted dlib CUDA build. Reproduce the GPU-memory and speed observations before determining whether the limitation is in video processing, CNN inference, or configuration. The issue does not define a specific code change or a clear completion test.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
computer-vision, machine-learning, performance
Issue type
Bug
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
18/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.