Skip to content
View sanikeit's full-sized avatar
🀯
🀯

Highlights

  • Pro

Block or report sanikeit

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.

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
sanikeit/README.md

Anikeit Sethi

Deep Learning Researcher Β· Generative AI Β· Computer Vision

MSc AI & ML @ TU Darmstadt Β |Β  BTech CSE @ IIT Indore

LinkedIn Google Scholar ResearchGate Portfolio Email


About Me

I'm an AI/ML researcher and engineer with a strong publication record, hands-on edge deployment experience, and a passion for building things that work in the real world β€” not just in notebooks.

My core expertise sits at the intersection of generative models, video understanding, and applied computer vision. I've taken systems from research prototype all the way to field deployment on resource-constrained hardware, and I enjoy working on problems where that full journey matters.

Currently pursuing my MSc at TU Darmstadt, where I'm also working as a research assistant on diffusion model steerability and federated learning for medical imaging.

πŸ… Institute Best Research Paper Award, 2025 β€” IIT Indore


What I Can Help With

Whether you're looking for a researcher, a collaborator, or a freelance engineer, here's where I add the most value:

Area What I Bring
πŸ”¬ Research Collaboration Novel model design, ablation studies, paper writing, benchmark evaluation
πŸ› οΈ Freelance / Contract Computer vision pipelines, model training & fine-tuning, edge deployment
πŸ’Ό Full-time / Internship Deep learning engineering, applied AI research, MLOps
πŸ₯ Medical AI Histology image analysis, federated learning, segmentation pipelines

Research Areas

Generative Models Novel GAN architectures for video understanding β€” CVAD-GAN, VALD-GAN, STemGAN, MAAD-GAN, iMAppGAN. Work spans anomaly detection, style transfer, and latent space manipulation.

Diffusion Model Steerability Training pipelines for text-to-image control; integrating Sparse Autoencoders as adapters for fine-grained generation steering.

Federated Learning for Medical Imaging Multi-client FL simulation on hyperspectral histology data (PaNCK & H&E staining); knowledge distillation / teacher-student extensions.

Edge AI & Real-world Deployment Optimized deep learning models for NVIDIA Jetson Nano (32 FPS real-time anomaly detection). Full stack from dataset collection to field testing.


Selected Publications

# Paper Venue Year
J.1 Attention-guided dual discriminator GAN for real-time video anomaly detection ⭐ Engineering Applications of AI 2024
J.2 CVAD-GAN: Constrained video anomaly detection via GAN Image and Vision Computing 2024
J.3 VALD-GAN: Video anomaly detection via latent discriminator augmented GAN Signal, Image and Video Processing 2024
J.4 STemGAN: Spatio-temporal GAN for video anomaly detection Applied Intelligence 2023
C.1 PAIN-GAN: Dual-stream encoder GAN for image style transfer ICONIP 2024 2025
C.2 iMAppGAN: Integrated motion-appearance GAN for anomaly detection ICONIP 2024 2025
C.3 Constraining latent space for unsupervised image-to-image translation CVIP 2024 2024
C.4 VALT GAN: Anomaly detection via latent space mining IEEE SSCI 2023 2023
C.5 MAAD-GAN: Memory-augmented attention discriminator GAN CVIP 2023 2023

⭐ Best Research Paper Award, IIT Indore 2025

πŸ“„ Full list on Google Scholar


Tech Stack

Core ML/DL

Python PyTorch TensorFlow Keras OpenCV scikit-learn

Data & Tooling

NumPy Pandas Matplotlib Docker Git

Languages & Web

C++ MATLAB TypeScript

Web3 & Blockchain

Solidity Ethereum Web3.js

Backend & Infrastructure

Next.js Fastify PostgreSQL Redis Supabase Vercel


Experience Snapshot

Hilfswissenschaftler β€” MEC Lab, TU Darmstadt (Jul–Dec 2025) Federated learning for medical hyperspectral imaging; knowledge distillation pipelines.

Hilfswissenschaftler β€” AIML Lab, TU Darmstadt (Jan–Jun 2025) Diffusion model steerability; Sparse Autoencoder integration for text-to-image control.

Translational Research Fellow β€” IIT Indore (Oct 2023–Aug 2024) End-to-end development and field deployment of highway surveillance AI system in collaboration with DevAditya Technocrats LLP and CSIR CEERI Pilani.

Research Intern β€” Computational Intelligence Lab, IIT Indore (Jan 2022–Sept 2023) GAN-based video anomaly detection research leading to multiple journal publications.


Freelance services I offer:

  • πŸ”— Smart contract development (ERC-20, vesting, presale mechanics)
  • 🏦 Token distribution platforms & investor portals
  • πŸ” Blockchain transaction scanning & on-chain reconciliation
  • 🌐 Web3 frontend integration (wallet connect, contract interaction)
  • πŸ€– Computer vision & deep learning pipelines
  • πŸš€ ML model deployment & optimization (including edge devices)

Open To

  • 🀝 Research collaborations in generative AI, video understanding, or medical imaging
  • πŸ’» Freelance projects in computer vision, model fine-tuning, or ML pipeline development
  • πŸ’Ό Full-time / internship roles in applied AI research or deep learning engineering
  • πŸ“§ Reach me at anikeitsethi@gmail.com or connect on LinkedIn

Based in Darmstadt, Germany Β· Open to remote & on-site opportunities

Pinned Loading

  1. Zenmap Zenmap Public

    Zenmap (GUI)

    1 1

  2. cell-segmentation cell-segmentation Public

    Python

  3. K-means-for-Evolving-Datastream K-means-for-Evolving-Datastream Public

    Jupyter Notebook

  4. G12project/FoodKart G12project/FoodKart Public

    Food Aggregator and Delivery Web app

    Python 1 4

  5. Rituraj-ksi/CVAD-GAN Rituraj-ksi/CVAD-GAN Public

    Python 2 1