🔤 OCR Neural Network
1600 →
256→128
(ReLU×2) →
36
classes (softmax)
Classes
All 62 — digits · A–Z · a–z
Alphanumeric, case-insensitive (36)
Letters, case-insensitive (26)
Uppercase A–Z (26)
Lowercase a–z (26)
Digits 0–9 (10) — easiest
H1
H2
LR
Epochs
augment ×
① Build dataset & train
■ Stop
Input — draw or generate a 40×40 glyph
Brush
erase
Clear
Or pick a glyph
char
font
Render → grid
Test fonts (marked ⚑) were
not
used in training — they show generalisation to unseen typefaces.
What the network actually sees (after centre + scale normalisation):
Prediction
—
Training
Click “Build dataset & train”. Needs internet access to fetch the Google Fonts.
Epoch
—
Train accuracy
(10 fonts seen, mixed styles)
—
Test accuracy
(6 unseen fonts ⚑, mixed styles)
—
Loss
—
Training data:
0
glyphs · Test data:
0
glyphs
Hidden-layer activations (current input)
Hidden layer 1
Hidden layer 2
Each bar is one neuron’s ReLU output — the network’s learned feature response to the glyph above (first 96 neurons of each layer shown).
Output layer —
36
classes (brightness = probability)
The yellow-outlined cell is the network’s pick.