Publications

MCP4IFC: IFC-Based Building Design Using Large Language Models
Bharathi Kannan Nithyanantham, Tobias Sesterhenn, Ashwin Nedungadi, Sergio Peral Garijo, Janis Zenkner, Christian Bartelt, Stefan Lüdtke
arXiv Preprint, 2025

An open-source framework enabling large language models to create, edit, and query IFC building models through the Model Context Protocol. Combines predefined BIM tools with dynamic code generation using retrieval-augmented generation, allowing LLMs to perform complex architectural design tasks via natural language instructions without direct BIM software interaction.

paper / code / project

Depth Fusion in Transformer-Based Video Object Detection
Bharathi Kannan Nithyanantham
Master's Thesis, University of Siegen, 2024

Investigated multiple depth-aware fusion strategies (early, mid, and late fusion) integrated into transformer-based architectures (DETR, Deformable-DETR) for both image-level and video-level object detection. Evaluated on SUN RGB-D and custom video benchmarks, demonstrating improved robustness under occlusion, motion blur, and scene clutter compared to RGB-only baselines.

code