For a project focusing on computer vision,I feel PyTorch is often recommended due to its ease of use, flexibility, and the strong support. it has within the research community. It’s particularly advantageous if you need to experiment with cutting-edge techniques or require dynamic model architectures. PyTorch also offers many pre-trained models and libraries specifically for computer vision tasks, such as torchvision and support GPU and CPU architectures.
Your correct,but I am finding difficult to get to an Impactful idea where I can use gen AI and do a startup I am even scared because these big giants are smooth in copying these things so how would I defend that one?
I think the thing to focus on is solving the customers problem rather than starting with “how do I use AI”. You may end up not using AI or just for a single feature of a bigger product.
Your right,I agree with it but,the thing is for a good idea to get example Chatgpt it was Google's transformers architecture where he had done that but still the idea of using was by openAI so to find such good ideas or trying to solve a problem it's kind of hard in this generation. What are your thoughts on this?
I am really not thinking about this but the thing is whenever I think of an idea and share that idea to someone they just say to my face that this will not work especially family and friends.
I had an idea to use AI with a phone's camera to identify any medicine, whether it's in a package or not. This AI would provide information about the medicine, including its uses, symptoms it treats, side effects, and the correct dosage based on the individual's health condition. The inspiration for this idea came from a personal experience in India, where some doctors prescribe higher dosages to earn commissions. My father was given medicine intended for heart or liver transplants, even though he had not undergone any transplant. This AI tool could help prevent such situations by providing accurate and reliable information about medications.
Sounds like a super-worthwhile, valuable & lives-enhancing (or even saving in some cases perhaps!) idea to me.
Your lesson here then: Dont ask people to judge your ideas, unless they're themselves full of good ideas often (in your estimate). Whether family or friends or anyone is irrelevant here on principle.
Here's another ai-but-not-gen-ai idea: app to see how one would look with some specific haircut, from multiple angles or ideally from all angles (feed enough photos for a good-enough "3d skull detection"). One could show their haircut choice rendered in-app to their barbers, and/or barbers could also subscribe to such an app to offer the feature to their present customer, perhaps by live-filming their head for a few seconds. (Something I'd like to exist but not badly enough to make it happen, compared to my other pursuits.)
