#visionlanguage resultados da pesquisa

HoneyBee: A 2.5M-sample vision-language reasoning dataset boosts VL model accuracy up to +38.9%, with smarter data curation (context, CoT, scaling) and 73% lower decoding cost. #AI #VisionLanguage

Encord’s EMM1 tests AI on real-world vision and language tasks. It reveals strengths and gaps in context understanding while pushing models to generalize better. Complexity and resources remain a challenge. #AI #Data #VisionLanguage encord.com/multimodal-dat…

Back from the break with Phillip Isola @phillip_isola on “On the Perceptual Distance Between Images and Text.” A fascinating and interactive look at how models (and humans!) measure similarity 👏🏻 #HiCV2025 #ICCV2025 #VisionLanguage

Join us as we explore how to build trustworthy, robust, and safe vision-language generative models (e.g., Text-to-Image & Image-to-Text models). #ICCV2025 #GenerativeAI #VisionLanguage #AITrustworthiness #ResponsibleAI

🚀 Temporal Understanding has been a missing piece of puzzle for multimodal large language models. 🧵 0/n Proud to announce our work, TimeWarp, a novel synthetic temporal preference data generation pipeline. #VideoLLMs #VisionLanguage #Genai #LLMs #TemporalUnderstanding

The special contribution? Our dataset goes beyond perception: we curated cultural reasoning questions that require knowledge to answer, not just visual cues. Check it out here 👉 seeingculture-benchmark.github.io #EMNLP #visionlanguage #vlm #benchmark #computervision

[1/4] We tested InternVL3.5-aligned GPT-OSS on MMMU and found visual alignment gaps: misreads, shallow grounding, and repetition. It’s an early step, not the solution—good scaffolding, weak seeing. #Multimodal #VisionLanguage

❓Docs read by models—but can they prove it? // RAD² X ensures traceable, agency-first choices in VLM pipelines. glcnd.io/transforming-d… #GLCND #RAD2X #VisionLanguage #ExplainableAI

Thrilled to see ERNIE 4.5 VL earn a top spot on the latest SuperCLUE-VLM benchmark. Time to build something new with vision!👀 #AI #Multimodal #VisionLanguage #LLM #Benchmark #ERNIE

Discover FastVLM, a breakthrough in Vision Language Models that boosts image resolution and speeds up processing, making text-rich image understanding more efficient! 🚀📸 #AI #VisionLanguage #TechInnovation rpst.cc/oMa5DZ

🚨 New model alert! 🚨 We've added OpenGVLab InternVL3_5-2B! It's a vision-language model. Get it running in LocalAI with: `local-ai run opengvlab_internvl3_5-2b` 😉 #LocalAI #VisionLanguage #NewModel

🎙️ AI Frontiers: Vision-Language Integration Breakthroughs (Aug 17, 2025) #AIFrontiers #VisionLanguage #MultimodalAI Watch full episode: youtube.com/watch?v=i08Hly…

Explore now → github.com/rednote-hilab/… #OCR #VisionLanguage #AI #opensource #MachineLearning

Great to see NVIDIA releasing a huge vision-language dataset with 3M samples! Open datasets like this will help accelerate research in OCR, VQA and captioning tasks. Excited to see what developers build with it. #AI #VisionLanguage

MiniGPT-4: Enhancing vision-language understanding using advanced large language models for multi-task learning. #AI #VisionLanguage #LLM #MachineLearning

💡 Dynamic Token Compression in DeepSeek-VL: ✂️ 40% fewer tokens for high-res images 📊 +3.1% accuracy on TextVQA Code: github.com/deepseek-ai/vl… #VisionLanguage @PublicAI_ @PublicAIData @DataBabies333

🚨 @ICCVConference 2025🚨 Happy to share that our paper: Visual Modality Prompt for Adapting Vision-Language Object Detectors was accepted at #ICCV2025 📄 Paper : arxiv.org/abs/2412.00622 💻 Code: github.com/heitorrapela/M… #ComputerVision #VisionLanguage #ObjectDetection #VLMs

TIMING MUST BE for it to last forever #ViSiONLANGUAGE #ObedienceOverSacrifice

MY FIRST STEP IS TO PLANT MY SEEDS WHEREVER I CAN PLANT THEM #ViSiONLANGUAGE LIVING 4EVA IS DA GOAL #2000NOW

@StemandThorn Love and Respect your work...... #ViSiONLANGUAGE

Discover FastVLM, a breakthrough in Vision Language Models that boosts image resolution and speeds up processing, making text-rich image understanding more efficient! 🚀📸 #AI #VisionLanguage #TechInnovation rpst.cc/oMa5DZ

You see me because eye #StayLit #NeverNeglect #ViSiONLANGUAGE 👤💡

📢Join us for our #ImageInWords poster presentation on Nov 12th, 11:00-12:30pm at #EMNLP . We'll be diving deep into hyper-detailed image descriptions and the impact on #VisionLanguage models. See you there!👋 #NLProc #ComputerVision @emnlpmeeting @AndreaDBurns @GoogleDeepMind 🧵

Back from the break with Phillip Isola @phillip_isola on “On the Perceptual Distance Between Images and Text.” A fascinating and interactive look at how models (and humans!) measure similarity 👏🏻 #HiCV2025 #ICCV2025 #VisionLanguage

HELLO WORLD........ #4500FILMS & #ViSiONLANGUAGE #MUUG #MagicalUnionUnderGOD @nuthouseradio L.A.G. COMING SOON!!!!

[1/4] We tested InternVL3.5-aligned GPT-OSS on MMMU and found visual alignment gaps: misreads, shallow grounding, and repetition. It’s an early step, not the solution—good scaffolding, weak seeing. #Multimodal #VisionLanguage

In 2023, researchers launched VisIT-Bench, a benchmark with 592 vision-language tasks spanning 70 categories like plot analysis and art knowledge. #AI #VisionLanguage @Stanford

🚀 Thrilled to announce our paper "TG-LLaVA: Text Guided LLaVA" accepted by @AAAI! We enhance vision encoders with text guidance, boosting performance without extra data. Huge thanks to the team! #AI #VisionLanguage

Thrilled to see ERNIE 4.5 VL earn a top spot on the latest SuperCLUE-VLM benchmark. Time to build something new with vision!👀 #AI #Multimodal #VisionLanguage #LLM #Benchmark #ERNIE

1/5 🖼️📚Introducing a new approach for vision-language pre-training: "Efficient Vision-Language Pre-training by Cluster Masking." This method enhances visual-language contrastive learning with a novel masking technique. #AI #MachineLearning #VisionLanguage

1/5 🌐"On the Test-Time Zero-Shot Generalization of Vision-Language Models: Do We Really Need Prompt Learning?" This paper explores the necessity of prompt learning for VLMs. #AI #MachineLearning #VisionLanguage

1/5 🗣️"MMDU: A Multi-Turn Multi-Image Dialog Understanding Benchmark and Instruction-Tuning Dataset for LVLMs." This benchmark evaluates LVLMs in complex dialog scenarios involving multiple images. #AI #MachineLearning #VisionLanguage

🔍 𝗪𝗵𝗮𝘁’𝘀 𝗻𝗲𝘄? We propose a new approach to fine-tuning large Vision-Language Models (VLMs) on resource-constrained clients in Federated Learning—essential for healthcare, where privacy matters most. #FederatedLearning #VisionLanguage #FoundationModels

🚨 @ICCVConference 2025🚨 Happy to share that our paper: Visual Modality Prompt for Adapting Vision-Language Object Detectors was accepted at #ICCV2025 📄 Paper : arxiv.org/abs/2412.00622 💻 Code: github.com/heitorrapela/M… #ComputerVision #VisionLanguage #ObjectDetection #VLMs

🔍 Today's top pick from @powerdrill Research Digest: 'Training-Free Unsupervised Prompt for Vision-Language Models'. Check out the link for a summary: app.powerdrill.ai/s/1jB88R #AI #VisionLanguage #Research

AI Model Unlocks a New Level of Image-Text Understanding 🔍📷✨ azoai.com/news/20241103/… #AI #MachineLearning #VisionLanguage #ImageText #ProLIP #ZeroShot #UncertaintyModeling #Innovation #DeepLearning #FutureTech @NAVER_AI_Lab

🔥 Discover the fascinating world of Multimodal Foundation Models, with the journey "From Specialists to General-Purpose Assistants" 🌐 Dive into the evolution of large models in #ComputerVision & #VisionLanguage! Paper: arxiv.org/abs/2309.10020 Tutorial: vlp-tutorial.github.io/2023/