In addition to this question: Is there a open-source pretrained TensorFlow Lite model for failed print and spaghetti detection? And in preparation of training a lightweight (TensorFlow Lite) model myself, perhaps labeling things with Meta's Segment Anything (SAM). Is there an open, perhaps crowd-sourced dataset of visual (image or video) data of failed and succesful prints?
Similar to, for example Mozilla Common Voice for voice recognition datasets, Common Crawl for natural language processing (NLP), and Open Images for computer vision (CV).
I would like to contribute to this dataset as well as use it for training an open source model that can run on efficient TPU's like Coral AI for Raspberry Pi, for the purpose of Spaghetti detection, empty bed detection and perhaps for other print defects.
