Florence-2 is a novel vision foundation model with a unified, prompt-based representation for a variety of computer vision and vision-language tasks.

VLM driven tool that processes surveillance videos, extracts frames, and generates insightful annotations using a fine-tuned Florence-2 Vision-Language Model. Includes a Gradio-based interface for querying and analyzing video footage.

Run SOTA Vision-Language Model Florence-2 on your data!

ViLMA (Vision-Language Model Active Monitoring) - A real-time desktop monitoring tool leveraging Florence-2

This application utilizes the powerful Florence-2 vision-language model from Microsoft to generate comprehensive captions for images. The model is capable of understanding visual content and expressing it in natural language.

The MultiModal-Vision-Language-Model-Training repository provides scripts for fine-tuning vision-language models (PaliGemma, BLIP-2, BLIP, SmolVLM, Qwen-VL, Florence-2) on SkinCAP and ROCOv2 datasets for medical image captioning. Optimized with LoRA and 4-bit quantization, it includes efficient training, evaluation (loss, accuracy, ROUGE, BLEU)

Florence-2 is a large vision-language model capable of various image and text generation tasks, such as object detection, captioning, and grounding. This demo allows users to interact with these capabilities by uploading images and selecting from various tasks.

No description available

A Multimodal Image Captioning and Audio Narration System Using the Florence 2 Vision Language Model.

Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks Model Summary This Hub repository contains a HuggingFace's transformers implementation of Florence-2 model from Microsoft. Florence-2 is an advanced vision foundation model that uses a prompt-based approach to handle a wide range of vision and vision-language tasks.

A comparative analysis of Vision-Language Models (Florence-2 vs. BLIP-2) performing dense region captioning on isolated objects segmented by SAM3.

This project implements an advanced generative AI pipeline for extracting and rating features from images. It combines the power of Florence-2, a state-of-the-art vision-language model, with a fine-tuned version of Mistral-v3, a cutting-edge large language model.

Smart Model Loader for ComfyUI — for vision-language models, text LLMs, and WD14 taggers across 8 backends (Transformers, GGUF, vLLM, SGLang, Ollama, llama.cpp, YOLO, WD14). Supports QwenVL, Mistral3, Florence-2, LLaVA, YOLO with multi-task chaining, few-shot training, and auto-download. V3 API + Nodes 2.0 compatible. NVIDIA/AMD/ROCm.