Skip to content
View Latchan-Ch's full-sized avatar
63 followers · 62 following

Organizations

@HaloMind-Research

Block or report Latchan-Ch

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Latchan-Ch/README.md

Hi, I'm Latchan Chhetri

About Me

Founder & Lead Research Architect @ Halo Mind Research Group

I engineer hardware-efficient neural mechanisms, specifically sub-quadratic State Space Models (Mamba) and more.

All of my active research and production codebases are hosted under the Halo Mind Research Organization.

(See my pinned repositories below for active/published works).

Core Architectures & Pipelines

  • Medical Vision: Geometric Residual Learning for 3D Segmentation (DDL-Mamba), Attention-Enhanced Swin Transformers, more..
  • Space AI: CNN-DDL frameworks for Space AI, Marginal Split Conformal Prediction wrappers, EO, Foundational Models,more..
  • Doc Analytics & OCR: Self-supervised OCR pipelines for degraded Ashokan Brahmi scripts, more..

Reach me at: latchanchhetri19@gmail.com

Personal Portfolio: latchan-ch.github.io

Research Group Page: halomind-research.github.io

Pinned Loading

  1. HaloMind-Research/DDV-GNet-Space HaloMind-Research/DDV-GNet-Space Public

    Official PyTorch implementation of DDV-GNet: A real-time (853 FPS) gated convolutional neural network for real-time aerospace defect detection.

    Jupyter Notebook 2

  2. Robust-Swin-Brain-Tumor Robust-Swin-Brain-Tumor Public

    Official implementation of "Robustness of Attention-Enhanced Swin Transformers Against Data Leakage in Brain Tumor Classification". SOTA accuracy (96.82%) on strict patient-level split.

    Jupyter Notebook 2

  3. HaloMind-Research/DDV-Mamba_for_Efficient_Remote_Sensing HaloMind-Research/DDV-Mamba_for_Efficient_Remote_Sensing Public

    DDV-Mamba: A highly efficient hierarchical vision model utilizing 2D Deep Delta Learning and SSM-Gated Aggregation for real-time satellite land-cover classification.

    Jupyter Notebook 3

  4. HaloMind-Research/SafeMed-SSL HaloMind-Research/SafeMed-SSL Public

    Code accompanying the bioRxiv preprint “Trust, but Verify”: uncertainty-gated semi-supervised malaria cell classification using Monte Carlo dropout and confidence-weighted pseudo-labeling.

    Jupyter Notebook 2

  5. HaloMind-Research/Interpretable-Solar-Defect-Detection HaloMind-Research/Interpretable-Solar-Defect-Detection Public

    Official code implementation for "Interpretable Solar Panel Defect Detection" . Uses Swin Transformers and Fuzzy Logic for Explainable AI (XAI).

    Jupyter Notebook 2

  6. HaloMind-Research/Optimizing-Deep-Learning-for-Brain-Tumor-Classification HaloMind-Research/Optimizing-Deep-Learning-for-Brain-Tumor-Classification Public

    Official implementation of the paper Optimizing Deep Learning for Brain Tumor Classification: A Comparative Ablation Study of Preprocessing and Augmentation Strategies. Includes patient-level data …

    Jupyter Notebook 2