karuvanan/MiniMax-H3-Director-Cut-Studio
Premiere-inspired PySide6 director studio for MiniMax H3 Ref2VA with AI shot planning, semantic media enrichment, timeline prompt reconciliation and shot-aware long-video rendering through ComfyUI.
5,366,387 AI resources indexedWorking toward the world’s largest AI resource index.
AI software for creative and professional work.
Local Inference Engines RAG and Retrieval AI Evaluation Tools AI Monitoring
Premiere-inspired PySide6 director studio for MiniMax H3 Ref2VA with AI shot planning, semantic media enrichment, timeline prompt reconciliation and shot-aware long-video rendering through ComfyUI.
BrainAgent — brain-inspired plugin for DeepSeek Harness (dsh): a pipeline of heuristic filters, statistical memory stores and context injections; reward-ledger + UCB1 bandit learning loop (RL-lite). 915 tests. Free noncommercial use.
Empirical proof of SOTA LLM (GPT-5/Gemini-Pro/Claude-Pro) context saturation in complex engineering. Contains the "Misuraca Protocol" for deterministic logical segmentation to prevent entropy drift.
Blip AI Agent HTTP Actions.
High-performance, offline document parser and data prep library for .NET RAG and AI pipelines. Extracts text and metadata (Word, Excel, PowerPoint, HTML, RTF, Email, EPUB, OpenDocument) and redacts PII and secrets locally before sending to LLMs. Features parallel streaming, incremental scans via manifests, pluggable format extractors, and built-in chunking.
ImageInsightAI: An image captioning package using BLIP model
Vision-language hallucination measured against a hand-verified adversarial set. BLIP hallucinates 4.5% but misses 29% of present objects; CLIP verification cuts hallucination 33-40% at a real cost in recall.
Implementation of Qformer from BLIP2 in Zeta Lego blocks.
Chain together LLMs for reasoning & orchestrate multiple large models for accomplishing complex tasks
CLI tool for image caption using BLIP & BLIP2
"An adaptable, AI-powered Disaster VQA system that performs real-time visual assessment of disaster imagery. By defaulting to the highly efficient BLIP model for rapid inference—while maintaining full architectural support for massive Vision-Language Models like LLaVA/Qwen-VL—it delivers contextual answers and confidence scores to accelerate emerge
Open-source writing skills for Claude Code, Codex, OpenCode, Antigravity, Gemini, OpenClaw, and Hermes. Your creativity. Your agent. Your book.
Try a broader search or clear some filters.
Platform labels describe intended use. Read each resource's author credits, requirements and available access methods.