boltgolt/howdy

[Feature Request] Continious retraining

Open

#342 opened on Apr 29, 2020

View on GitHub
 (4 comments) (5 reactions) (0 assignees)Python (4,836 stars) (268 forks)batch import
enhancementhelp wanted

Description

Humans often change overtime. Hair style changes, clothing changes, bruises and cuts appear and skin wrinkles.

The issue i have had with howdy is that my beard throws the detection off.

Is it possible that, if it detects a well matching face, that howdy will remake the model with the addition of the new data. So that howdy will be able to grow a "life cycle" of your face, such as in my case the growth and removal of my beard

Contributor guide

Tech stack
python
Domain
securityauthenticationmachine learning
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-2 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
mostly clear
Prerequisites
Pythonbasic face recognition concepts
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
30
Research direction
Investigate the current model training pipeline in howdy (likely in src/ or models/ directory). Consider approaches like incremental learning or periodic retraining with new face captures. Review any existing discussions in the issue comments for constraints. Examine the dlib or face recognition library usage to understand model update feasibility.