Tiny CNN Simulator (32×32, Binary)

LR Epochs Batch

Training Monitor

Dataset size: 0

Epoch: 0

Loss: -

Acc: -

Training Images

Class 1 (mapped to label 0)

Class 2 (mapped to label 1)

Uploads are auto-resized to 32×32 (grayscale) and added to the training dataset.

Conv Filters (Realtime)

Prediction

-

Feature Maps (Conv Layer)

Notes

Binary classifier with tiny CNN: Conv→ReLU→Flatten→Dense→ReLU→Dense→Softmax
Demo loader + per-class uploads supported
Filters and feature maps visualized in realtime