https://github.com/IVGSZ/Flash-VStream
공식 코드Flash-VStream: Efficient Real-Time Understanding for Long Video Streams
Papers with Code 아카이브
MME
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
- 하위 태스크
- 0
- 평가표
- 0
- 보존된 평가 기록
- 0
MME is a comprehensive evaluation benchmark for multimodal large language models. It measures both perception and cognition abilities on a total of 14 subtasks, including existence, count, position, color, poster, celebrity, scene, landmark, artwork, OCR, commonsense reasoning, numerical calculation, text translation, and code reasoning.
코드
관련 논문과 평가 결과에 연결된 구현입니다. 출처에서 확인된 저자 구현에는 공식 코드 표시가 있습니다.
https://github.com/InternLM/InternLM-XComposer/tree/main/projects/ShareGPT4V
공식 코드ShareGPT4V: Improving Large Multi-Modal Models with Better Captions
https://github.com/NVlabs/FRAG
공식 코드FRAG: Frame Selection Augmented Generation for Long Video and Long Document Understanding
https://github.com/Ziyang412/VideoTree
공식 코드VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos
https://github.com/bradyfu/awesome-multimodal-large-language-models
공식 코드A Challenger to GPT-4V? Early Explorations of Gemini in Visual Expertise
https://github.com/bytedance/pargo
공식 코드MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models
https://github.com/ceezh/silvr
공식 코드https://github.com/comeandcode/MRT
공식 코드Re-Imagining Multimodal Instruction Tuning: A Representation View
https://github.com/dcdmllm/cheetah
공식 코드Fine-tuning Multimodal LLMs to Follow Zero-shot Demonstrative Instructions
https://github.com/dcdmllm/hyperllava
공식 코드HyperLLaVA: Dynamic Visual and Language Expert Tuning for Multimodal Large Language Models
https://github.com/dingchenyang99/pensieve
공식 코드Pensieve: Retrospect-then-Compare Mitigates Visual Hallucination
https://github.com/dvlab-research/Lyra
공식 코드Lyra: An Efficient and Speech-Centric Framework for Omni-Cognition
https://github.com/dvlab-research/prompt-highlighter
공식 코드Prompt Highlighter: Interactive Control for Multi-Modal LLMs
https://github.com/eit-nlp/connector-selection-for-mllm
공식 코드https://github.com/evolvinglmms-lab/longva
공식 코드https://github.com/evolvinglmms-lab/mgpo
공식 코드High-Resolution Visual Reasoning via Multi-Turn Grounding-Based Reinforcement Learning
https://github.com/fdamken/decoding_multilingual_moral_preferences
공식 코드Decoding Multilingual Moral Preferences: Unveiling LLM's Biases Through the Moral Machine Experiment
https://github.com/haozhezhao/mic
공식 코드MMICL: Empowering Vision-language Model with Multi-Modal In-Context Learning
https://github.com/internlm/internlm-xcomposer
공식 코드https://github.com/jiangsongtao/vtprompt
공식 코드https://github.com/kakaobrain/honeybee
공식 코드https://github.com/lee-lab558/iava
공식 코드Instruction-Aligned Visual Attention for Mitigating Hallucinations in Large Vision-Language Models
https://github.com/leon1207/video-rag-master
공식 코드Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension
https://github.com/mac-automl/quota
공식 코드QuoTA: Query-oriented Token Assignment via CoT Query Decouple for Long Video Comprehension
https://github.com/mlpc-ucsd/bliva
공식 코드BLIVA: A Simple Multimodal LLM for Better Handling of Text-Rich Visual Questions
https://github.com/nlpcode/tubench
공식 코드TUBench: Benchmarking Large Vision-Language Models on Trustworthiness with Unanswerable Questions
https://github.com/opendatalab/ha-dpo
공식 코드Beyond Hallucinations: Enhancing LVLMs through Hallucination-Aware Direct Preference Optimization
https://github.com/ouyangkun10/spacer
공식 코드https://github.com/ouyangkun10/spatial-r1
공식 코드https://github.com/p1k0pan/ICD
공식 코드Mitigating Hallucinations in Large Vision-Language Models with Instruction Contrastive Decoding
https://github.com/peijin360/vsre
공식 코드https://github.com/pkunlp-icler/mic
공식 코드MMICL: Empowering Vision-language Model with Multi-Modal In-Context Learning
https://github.com/rucaibox/comvint
공식 코드https://github.com/sming256/bolt
공식 코드BOLT: Boost Large Vision-Language Model Without Training for Long-form Video Understanding
https://github.com/the-martyr/causalmm
공식 코드https://github.com/vec-ai/wikihow-tiir
공식 코드출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.