SVHN-Recognition

This project is a PyTorch implementation that uses deep CNN to recognize multi-digit numbers using the SVHN dataset derived from Google Street View house numbers, each picture contains a set of numbers from 0 to 9, the model is tested to have 89% accuracy.

Environment

conda create -n svhn --yes --file conda.txt
conda activate svhn
pip install -r requirements.txt

Usage

1. Clone the source code:

git clone git@github.com:MuGeminorum/SVHN-Recognition.git
cd SVHN-Recognition

2. Run convert_to_lmdb.py
3. Run train.py

Params

Steps	GPU	Batch Size	Learning Rate	Patience	Decay Step	Decay Rate	Accuracy
122000	GTX 1080 Ti	512	0.01	100	625	0.9	89.21%

Training curve

Reference

[1] Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks