Back to Projects
Deep Learning Image Recognition
AI & Machine LearningOngoing2024

Deep Learning Image Recognition

AI-powered image classification system designed for healthcare diagnostics and intelligent recognition tasks.

PythonTensorFlowKerasFastAPIDockerPostgreSQL

Overview

This project develops a deep learning pipeline using convolutional neural networks to classify medical images with high accuracy.

Key Features

1

Multi-class image classification using CNNs

2

DICOM medical image format support

3

Real-time inference with under 200ms response time

4

Explainability layer showing model confidence regions

5

Integration with hospital record systems

6

Continuous model retraining pipeline

Impact

Improved medical diagnosis speed by 40% through automated image analysis.

40%

Faster Diagnosis

95%

Model Accuracy

12

Hospitals Piloting

10K+

Images Processed

Team

Faculty Advisor

D

Dr. Almaz Haile

Students

A
Amir Tadesse
L
Liya Bekele
S
Samuel Girma
https://demo.com/