Implementation of Looking at Boundary paper

Job ID: 33587540

Budget: ₹12,500 – ₹37,500 INR

Looking for a implementation in python of the above mentioned paper with ResUTNet-50. More info can be found at this git: https://github.com/yhygao/UTNet

For the joint learning part, refer to this: https://github.com/jiangyao-scu/JL-DCF-pytorch

The main change the coder should make is use Split-and-Merge EMA instead of EMA for feature aggregation. Relevant papers are attached.

Also interested in demonstration of alternate squeeze-and-excitation blocks effect on performance but that's not a necessity. Please reach out only with confidence. The last person didn't deliver and cost me time.