machine-learning-image-processing-

Machine Learning – Infant Visual Recognition Analysis

https://github.com/danielco1111/machine-learning-image-processing-

Science Score: 26.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
    Found codemeta.json file
  • .zenodo.json file
    Found .zenodo.json file
  • DOI references
  • Academic publication links
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (9.5%) to scientific vocabulary
Last synced: 10 months ago · JSON representation

Repository

Machine Learning – Infant Visual Recognition Analysis

Basic Info
  • Host: GitHub
  • Owner: DanielCO1111
  • License: agpl-3.0
  • Language: Python
  • Default Branch: main
  • Size: 999 MB
Statistics
  • Stars: 0
  • Watchers: 0
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Created about 1 year ago · Last pushed about 1 year ago
Metadata Files
Readme Contributing License Citation

README.md

Infant Visual Recognition Analysis

Image Processing & Machine Learning Pipeline with YOLO, Python, and Pandas.

Overview

This project analyzes infant visual recognition behavior from video data using a fully automated image processing and machine learning pipeline. The pipeline combines video frame extraction, object detection with YOLOv5, feature engineering, and statistical comparison of experimental conditions (such as blurred vs. unblurred video). The codebase leverages Python, Pandas, NumPy, and OpenCV for robust, reproducible analysis.

Main Components

Frame Extraction:

Automated splitting of video files into frames for downstream analysis.

Object Detection (YOLOv5):

Using YOLOv5 to identify and locate relevant features in each frame (such as faces or specific visual targets).

Feature Engineering:

Extracting and aggregating key data points from detected objects (e.g., position, size, detection confidence).

Statistical Analysis:

Using Pandas and NumPy to compare visual behavior across different experimental setups (e.g., blurred vs. unblurred videos), generating clear visualizations and CSV reports.

Automation:

Scripts and notebooks for running the full pipeline in batch mode, supporting reproducibility for future experiments.

Visual Examples of the functionalities:


The image shows the output of a YOLO object detection model, which has identified and labeled a dog, a teddy bear, a potted plant, and a person (partially visible), each with a bounding box and a confidence score.

image


image image

In these two images, the top photo is clear (unblurred) and the bottom photo is blurred before running the object detection algorithm. This simulates the difference between adult vision (sharp) and infant vision (blurred). You can see that the detection confidence scores (the numbers above each box) are lower in the blurred image, showing that the algorithm finds it harder to recognize objects when the image is blurryjust like its more difficult for infants to recognize objects due to their limited vision.


image

image

Technologies

Python 3.x

YOLOv5 (Ultralytics, PyTorch-based)

OpenCV, Pillow

Pandas, NumPy

Matplotlib

Jupyter Notebooks

Owner

  • Name: Daniel Cohen
  • Login: DanielCO1111
  • Kind: user

GitHub Events

Total
  • Member event: 1
  • Push event: 14
  • Create event: 1
Last Year
  • Member event: 1
  • Push event: 14
  • Create event: 1