rohit-kundu / mfsnet

Based on our paper on skin lesion segmentation: "MFSNet: A Multi Focus Segmentation Network for Skin Lesion Segmentation"

Welcome to my GitHub Space!

💬 I am Rohit Kundu (he/him/his), a Doctor of Philosophy (Ph.D.) student at the Video Computing Group in the University of California, Riverside, USA working under the supervision of Prof. Amit K. Roy-Chowdhury. I work on low-supervision Deep Learning mainly for Vision-Language applications.

💬 In June 2022, I completed my Bachelors of Engineering (B.E.) degree from Jadavpur University, India, majoring in Electrical Engineering.

💬 Check out my Google Scholar profile for a complete list of my publications.

💬 ResearchGate | ORCiD

💬 For more details, visit my personal homepage: https://rohit-kundu.github.io/

⚡ News: ⚡

• June 2024: Joined Google (YouTube) as a Student Researcher.

Reach Me:

• rohit [dot] kundu [at] email [dot] ucr [dot] edu (Preferred)
• Twitter
• LinkedIn

mfsnet's Issues

poor preprocessing effect

Hi!
I'm having some problems with image preprocessing.The original configuration code provided does not have a good effect after preprocessing. Is there something wrong with me? Looking forward to your reply!
Thank you very much!

Best,
Quince.

(original)

(after preprocessing)

Weights for MFSNet

Hello, @Rohit-Kundu @hritam-98
Thanks for your paper! Could you provide weights for your model?

When loading the training model during the test, an error was reported that the model did not match

Hi！
The model MFSNet.pth saved by training does not match when it is loaded in the test, and then I find that some parts of train network and test network are different. May I ask how you apply the trained model to the test?Looking forward to your reply！Thank you very much!
Best,
Quince.