I would like to ask you for advice: I am completely new to this field (actually I came from NLP), and now playing with pet project where I want to classify camera movement
For example, here camera is moving around the "main" scene.
While here camera stands in the same place, while rotating to the right.
If I will use optical flow, I will get more or less the same output for both videos.
So, what are your thoughts about it? Does your model applicable to this case? What about generalization?
Hi, first, thanks for your work!)
I would like to ask you for advice: I am completely new to this field (actually I came from NLP), and now playing with pet project where I want to classify camera movement
For example, here camera is moving around the "main" scene.![Mar-26-2023+14-55-33](https://github.com/princeton-vl/DPVO/assets/45794881/734adf82-e70f-4c56-93dc-7ef0a4f15be3)
While here camera stands in the same place, while rotating to the right.![Apr-16-2023+21-12-39](https://github.com/princeton-vl/DPVO/assets/45794881/297f769d-c40a-43d9-83ae-f5b94df9d764)
If I will use optical flow, I will get more or less the same output for both videos.
So, what are your thoughts about it? Does your model applicable to this case? What about generalization?