Skip to content

AndersonYu7/FPGA_HandWrite_Recognition

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 

Repository files navigation

FPGA_HandWrite_Recognition

This project aims to implement a handwritten digit recognition model using FPGA (Field Programmable Gate Array) for hardware acceleration. The model recognizes digits from 0 to 9 and uses a simple three-layer deep neural network (DNN) for inference. This document details the project structure, model architecture, and usage instructions.

Model Architecture

The DNN model has a simple three-layer fully connected architecture as follows:

  1. Input Layer: 28 × 28 input features (handwritten digit images, flattened to 784 features).
  2. Fully Connected Layer 1: 128 neurons with ReLU activation.
    • Dropout1: Randomly drops 30% of neurons during training.
  3. Fully Connected Layer 2: 128 neurons with ReLU activation.
    • Dropout2: Randomly drops 30% of neurons during training.
  4. Output Layer: 10 neurons corresponding to digit classes (0-9) with Softmax activation.
  • Optimizer: Adam, learning rate = 0.001
  • Loss Function: Sparse Categorical Crossentropy

Hardware

  • DE2-70 FPGA development board
  • DE2-80 LTM Display

Result

756793522.776650.mp4

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published