pySLAM is a hybrid Python/C++ Visual SLAM pipeline supporting monocular, stereo, and RGB-D cameras. It provides a broad set of modern local and global feature extractors, multiple loop-closure strategies, a volumetric reconstruction module, integrated depth-prediction models, and semantic segmentation capabilities for enhanced scene understanding.

[CVPR 2023] Unsupervised Continual Semantic Adaptation through Neural Rendering

Knowledge Graph-enhanced RAG system for in-depth analysis of crime film narratives.

A new package would process textual descriptions of virtual reality scenes or environments and return structured, validated outputs that describe the scene in a standardized format. It would use an LL

Vectra is an API-first vector database designed for fast, meaningful semantic search. It enables applications to generate vectors, store embeddings, and perform high-performance similarity searches via simple HTTP requests, without adding computational load to core systems.

A comprehensive framework for fine-tuning and evaluating Large Language Models on semantic understanding tasks (WSD & WiC) with LoRA

Docker-based web demo for Annif using GermEval 2025 LLMs4Subjects models (BM ensemble, Bonsai, MLLM).

🌐 Streamline VR scene creation by converting natural-language descriptions into structured formats with vrscene-parser.

ProofCore is a browser-native, 100% offline-first, hybrid mathematical proof verification engine. It combines rigorous symbolic math with semantic understanding to reliably verify mathematical proofs, offering zero external dependencies and production-ready quality