Computer Vision vs Image Processing 

Thanks!

Dear Leo..thanks for the red heart = ) - I am researching about CV/ML& AI bcs. I have this little simple project idea that I beleive can help people struggling to make a livelihood day in day out and also donate funds to computer science research to help people with dyslexia, disgraphia, autisim, etc. I live in México and I would love to have some direction from someone expert in this field. Will that be possible that you can take a quick look of my project idea ? If possible, would you pls. email to mrosado.desart@gmail.com ? Thanks a lot Leo ! greetings from México City. =) Have a nice evening.

I think what he was suggesting about DNNs in that example was much more advanced than basic classification. With a shallow ML classifier, you can classify images into one of many well-defined categories. DNNs can do the same but with a couple of advantages: 1) depending on the type of data, a DNN can be much more accurate 2) DNNs can be trained to not only classify data, but describe it as well. Expanding on the second point, with a shallow classifier, you could bucket photos into ones that contain hotdogs and ones that don't. With a DNN, you could train the model to not only identify photos with hotdogs but also describe them with context (e.g. "a hotdog in a bun with mustard on top").

I have an M1 Macbook Pro. Different computer vision tasks may have different computational complexities. For some tasks, even a Raspberry Pi would be powerful enough. For others, you may need to use cloud GPUs.

Adobe Photoshop is a proprietary software that implements image processing operations. We don't know what they exactly use but you can get very close results with open source implementations in Python for many types of image processing operations.

@@leoisikdogan thank you very much for the clarification.

Thanks!

Thanks!