lightweight-model

CVPR 2026 (Main) | MambaLiteUNet: Cross-Gated Adaptive Feature Fusion for Robust Skin Lesion Segmentation

GCNfold: A novel lightweight model with valid extractors for RNA secondary structure prediction

AudioMuse-AI-DCLAP is a lightweight, high-speed distilled version of LAION CLAP, designed for fast and efficient text-to-music search

A minimal YOLOv8n-based object detection project using the lightweight Nano version of the model for fast and efficient training and inference on small datasets like coco128.

BMVC 2025 (Main) Oral (3%) | AULUNet: An Adaptive Ultra-Lightweight UNet Framework for Efficient Skin Lesion Segmentation in Resource-Constrained Environments

🔍 Enhance medical imaging with a lightweight CNN model that offers over 91% accuracy and integrated explainability for better clinical trust.

Official repository for paper "Enhancing 3D Point Cloud Classification with ModelNet-R and Point-SkipNet"

The official repository for "GIVTED-Net: GhostNet-Mobile Involution ViT Encoder-Decoder Network for Lightweight Medical Image Segmentation."

A lightweight Explainable AI CNN for PathMNIST medical imaging, achieving 91%+ accuracy with Integrated Gradients and SQLite-based attribution storage. Built in PyTorch, this scalable model delivers high performance, transparency, and real-world readiness, making it ideal for medical AI, edge deployment, and explainable deep learning research.

Rule-based NLP chatbot using TF-IDF & cosine similarity over 1000+ real-world QnA — fast, offline & LLM-free.

FMCNet: Enhancing Anime Image Super-Resolution through Frequency-Aware Multi-Scale Consistency

Lightweight Siamese Network for Remote Sensing Building Change Detection | 轻量化孪生网络遥感建筑物变化检测（LEVIR-CD 数据集）

🚀 Production-ready API for detecting toxic comments in Russian text, powered by a lightweight PyTorch CNN model (~5MB).

A lightweight Few-Shot Image Classifier built with PyTorch, capable of recognizing new classes from just a few examples using prototype-based learning. Designed for easy model export and integration.

LiteCDNet 公开仓库 | Public repo for training, evaluation, and ablation experiments

BSc thesis: energy and performance profiling of lightweight object detection models for autonomous vehicle perception. RAPL measurement scripts, raw results, and analysis.

the purest edition of Sphere. An Autonomous Metabolic Information Infrastructure for AI Agents Sphere is not a database. It is a circulation system — knowledge enters, gets evaluated by AI agents, crystallizes or fades, and moves on. Intelligence is never stored here; it passes through.

📄 Lightweight Q&A chatbot for PDF, DOCX, and TXT files. Built with Streamlit, LangChain, HuggingFace, and ChromaDB — no large LLMs required.

VisionicAI: Ultra-lightweight object detection model optimized for NPU. 1인 주도로 아키텍처 설계 및 학습 파이프라인 전 과정 개발.

Lightweight Separable CNN Model for Fire Image Classification