thawro/yolov8-digits-detection

Added instruction

Added canvas recording

refactored video uploading

Video uploading works good

Example video works niiiice

Swapped videos

Finished videos uploading

Added example video upload

Added image loading to sketch detector

Added Image based object detector

sketchCanvas finished

Added support for mobile drawing (however it still needs change because double tap is needed now)

Added Custom slider with nicer appearance

Added changable canvas size and iou/score thresholds sliders

Fixed bug with H > W image size

minor changes

Enabled choosing (only in code for now) between ONNX and not ONNX pre/post processing for frontend app

Added docstrings for visualization

Refactored code and moved object detectors to separate modules

React app uses all three onnx files (preprocessing, detection, nms)

Added preprocessing ONNX to .py file

Changed previous nms onnx file to the new one which returns boxes, class scores and class ids

Added onnx_nms file with function to create ONNX NMS

almost working full nms

working basic nms onnx in notebook

Removed thresholds from YoloOnnxModel init

removed unnecessary files

added user 1000 to dockerfile

added root to dockerfile

added root to dockerfile

minor updates

changed useradd to adduser in dockerfile

Changed user in dockerfile

Changed app port to 3000 (react)

Updated dockerfile

Moved react app to dockerfile

Added frontend in ReactJS

Added YOLO dataset generation from HWD+ digits dataset

Upaded visualisation to use the one from ultralytics

Added INFO.md

Added model file

Added src files

upd readme

Merge branch 'main' of https://huggingface.co/spaces/thawro/yolov8-digits-detection into main

removed readme

initial commit

added README