from js import handTrack, setTimeout, requestAnimationFrame, player from pyodide import create_once_callable import asyncio context = canvas.element.getContext("2d") isVideo = False model = None last_position = 0 direction = "stop" modelParams = { "flipHorizontal": True, # flip e.g for video "maxNumBoxes": 20, # maximum number of boxes to detect "iouThreshold": 0.5, # ioU threshold for non-max suppression "scoreThreshold": 0.6, # confidence threshold for predictions. } def toggle_video(evt): global isVideo player.jump() if (not isVideo): update_note.write("Starting video") pyscript.run_until_complete(start_video()) else: update_note.write("Stopping video") handTrack.stopVideo(video.element) isVideo = False update_note.write("Video stopped") async def start_video(): global isVideo update_note.write("Inside start video") status = await handTrack.startVideo(video.element) console.log("video started", status) if status: update_note.write("Video started. Now tracking") isVideo = True console.log( "Calling RUN DETECTION") y = await run_detection() else: update_note.write( "Please enable video") def sync_run_detection(evt): pyscript.run_until_complete(run_detection()) async def run_detection(): global model global isVideo global last_position global direction predictions = await model.detect(video.element) model.renderPredictions(predictions, canvas.element, context, video.element); if predictions: curr_position = predictions[0].bbox[0] + (predictions[0].bbox[2] / 2) delta = last_position - curr_position last_position = curr_position #console.log(delta, curr_position, last_position) if abs(delta) < 2: direction = "stop" elif delta > 0: direction = "left" else: direction = "right" for prediction in predictions: if prediction.label == 'open': player.jump() elif prediction.label == 'close': player.crouch() if (isVideo): await requestAnimationFrame(create_once_callable(sync_run_detection)); def handle_model(lmodel): global model model = lmodel update_note.write("Loaded Model!") async def start(): model = await handTrack.load(modelParams)#.then(handle_model) handle_model(model) pyscript.run_until_complete(start())

Use < > to move, ↓ to crouch and x to jump. If video is enabled, say hi to jump as well!

loading model ..