Osiyo. Dohiju? Hey. Welcome back.
The last few days I have made a ton of progress. Things that took me weeks to do with AngularJS and OpenCV4NodeJs have taken less than a couple days. It’s not the fault of the frameworks. They do what they’re supposed to. They’re just not for me. Granted some of the speech recognition work I had already done before and just rewrote so it would work with this.
The progress includes streaming video to a page in Flask. It can process an image in TesseractJs. It also processes voice commands and will speak back. With a voice command you can control aspects of the view (the page) and can turn on and off video effects. You can load a PDF. Though right now I am working on loading a PDF by voice. I’m also working on the ability to scroll using voice. It uses AJAX to process commands on the server.
On the list of first iteration items I’m adding are Tensorflow, Keras, finalize layout and separate templates, add multiple camera sources and display, take screen capture of video and process with Tesseract, set up individual overlays for each camera view (so one can detect motion, we’ll say, and another can recognize objects, or whatever, or they can both do the same thing), object recognition, face recognition, gestures, streaming OCR for multiple languages, and finally NLP/NLU for intents.
I know that is a lot to accomplish for the start. The first item I need to work on is fixing Tensorflow on my local. I keep having an issue with it running that it can’t find a library.
Until next time. Dodadagohvi.