Looking for a machine learning expert who will help with the following:
- Develop an algorithm that detects objects within a video (whether you use YOLO v3 or prefer a better option)
- We will give the video that will be used as a dataset but it need to be prepared and labeled properly.
- The algorithm will of course be trained on this dataset and tried on a video that we will also provide.
- The algorithm must be explained in brief to a member of our team who knows machine learning, so it will not be difficult and will be fast.
- The algorithm (pre-trained) must be able to be executed directly either on colab or any GPU.
- Transfer learning must be applied to increase the efficiency of detection.
- We want an algorithm that is real time that can directly detect objects in one second or less.
- The algorithm must also be self trained.
We are thinking of starting a prototype while using the already existing YOLO v3 with a another pre-trained model to increase YOLO's efficiency and this way the algorithm would be very easy to develop. We just need someone to help us implement it on windows with the above requirements.
About the recuiterMember since Sep 1, 2017 Isabella
from Fukuoka, Japan