I am a volunteer intern (funds are frozen for uni research assistants due to the pandemic) on a CV/ML project. Awhile back, I was assigned to detect and label moving objects in a video. After doing extensive research, I figured out that predicting, rather than only detecting their movement would be more precise and better in the long run. I have done tons of research and came up with an approach toward it, and I am fully committed to this now.
The project was assigned by one of my supervisor's PhD students and I haven't spoken to my actual supervisor all summer. Should I: (a) go ahead with the project and send him my final work when I come up with a viable solution and potentially the abstract for the paper, (b) email him and ask him if I can move forward with this approach.
Either way, the idea I have and this area is super interesting and has research potential. If he said no, I'd probably explore it independently anyways. Only labeling the data also isn't useful experience for me, as I want to go to grad school. I was told research assistants you shouldn't be afraid to advocate for what they want, and this is an area I am highly interested in - and I believe it will still benefit them.