AcaiOMR

An optical music recognition model for pianoform music

Upload an image to transcribe

Selected image:

Preview of the image to run inference on

Split Systems

Draw a box around each musical system in the image

Controls

Click and drag on the image to create a bounding box
Click a box to resize or delete it
Click and drag a box to move it
Boxes can overlap

Inference Settings

Model weights path (defined in acai_omr/config.py): tf_omr_train/vitomr.pth

Max inference token length

Inference progress:

Encoding image latent

Generating sequence

Complete lmx sequence:

Average confidence per token:

Transcribed score: