Im looking at training a tensorflow.js model for object detection within a graph, I'm currently using a coco-ssd model to detect objects with a webcam, but want to make it so that it now detects graph objects when a graph is shown to the camera, I couldn't find any pre-trained models that were specific to graph objects eg. x and y axis, bar, line, labels etc. Does anyone know how to train a model? I know I will need training data ie. images of graphs and so on but what else would I need and how would I go about adding it to the code I currently use
question from:
https://stackoverflow.com/questions/66063163/training-a-tensorflow-js-model 与恶龙缠斗过久,自身亦成为恶龙；凝视深渊过久,深渊将回以凝视…
