Tagged articles
VLM
17 articles · Page 1 of 1
AgentVLN introduces a VLM-as-Brain architecture for vision-language navigation, using cross-space representation mapping to translate 3D paths into 2D visual prompts, context-driven self-correction for error recovery, and query-driven perceptual chain-of-thought for active information gathering, achieving state-of-the-art results on R2R-CE and RxR-CE benchmarks with a 3B model deployable on Jetson edge devices.
AgentVLNCross-Space Representation MappingECCV 2026
0 likes · 10 min read

JD’s Oxygen AIIC combines human‑in‑the‑loop ontology engineering, a semantic search‑then‑discrimination pipeline, and a self‑evolving multi‑task LLM/VLM model to produce high‑quality product knowledge for over a hundred thousand categories and billions of daily SKU updates, boosting search coverage to 80%, attribute auto‑fill to over 80%, cutting quality issues by 37% and raising click‑through by 9% while achieving 94.2% precision and 82.8% recall.
JD.comLLMSemantic Search
0 likes · 21 min read

The DOPD paper introduces an advantage‑aware dual distillation framework that eliminates the privilege illusion, dynamically selects token‑wise strategies, and delivers up to 7.5‑point gains on LLM benchmarks while closing 89.8% of the teacher‑student gap and showing strong robustness across model sizes.
LLMModel ScalingVLM
0 likes · 9 min read

The article analyzes the limitations of traditional UI automation, introduces the AI‑Native ai_uitester pipeline that converts test‑case data with LLM enhancement, implements AI‑driven debugging and self‑healing, and unifies cross‑platform execution through a VLM‑based engine, backed by real‑world metrics.
AI TestingLLMPrompt engineering
0 likes · 17 min read

Doubao's Model 2.1, unveiled at the Force conference, pushes daily token usage past 180 trillion, captures 49.5% of China's public‑cloud MaaS market, tops code and agent benchmarks, delivers repository‑level coding, advanced multi‑modal reasoning, and introduces cost‑effective Pro and Turbo variants with a new Deep Think inference mode.
AI benchmarkingDoubaoLLM
0 likes · 11 min read

Traditional pixel‑by‑pixel UI comparison breaks on complex CAD drawings due to semantic changes, so a team built a visual‑language‑model fine‑tuning pipeline that turns failure cases into training data, achieves ~95% AI accuracy, improves regression efficiency by over 40%, and now powers hundreds of daily automation tests.
AI monitoringModel EvaluationUI automation
0 likes · 12 min read

UniRL is an open‑source, distributed reinforcement‑learning post‑training framework that consolidates fragmented pipelines for image, video, and language‑vision models, offering a unified rollout‑reward‑advantage‑train‑sync contract, extensive model support, built‑in algorithms, and multi‑modal reward components to lower engineering barriers in AIGC research.
LLMMultimodal RLReward Engineering
0 likes · 10 min read

Meta’s VLM³ demonstrates that a plain vision‑language model, when trained on large‑scale data with simple camera‑focal‑length and pixel‑space normalization, matches or surpasses expert 3D vision models across monocular depth estimation, object‑level understanding, pixel‑matching and camera‑pose tasks, eliminating the need for task‑specific architectures, loss functions, data augmentations or regression formulations.
3D visionMetaScaling Laws
0 likes · 6 min read

Researchers from ETH Zurich show that tiny, human‑imperceptible perturbations to a single image can fool leading visual language models—including GPT‑5.4, Claude Opus 4.6, and Grok—into confidently delivering fabricated answers, enabling misinformation amplification, defamation, content‑filter evasion, and large‑scale AI authority laundering.
AI safetyClaude OpusGPT-5.4
0 likes · 7 min read

The article examines the intrinsic challenges of extracting structured text from PDFs for Retrieval‑Augmented Generation—such as missing reading order, table reconstruction, font encoding, and scanned images—and compares lightweight libraries, AI‑enhanced frameworks, commercial APIs, and visual language models as practical solutions.
AI FrameworksOCRPDF parsing
0 likes · 23 min read

PaddleOCR‑VL‑1.5, the latest Baidu release, uses only 0.9 B parameters to achieve 94.5% accuracy on OmniDocBench v1.5, surpassing larger open‑source and commercial OCR models, while offering multi‑task, multi‑language support, lightweight deployment, and detailed performance benchmarks.
DeepSeek-OCRGPU InferenceMulti-language
0 likes · 9 min read

VLM‑FO1 introduces a generate‑plus‑reference paradigm that replaces coordinate generation with region token referencing, adding plug‑in modules such as a proposal generator, a hybrid fine‑grained encoder, and a region‑language connector to give any pretrained visual language model accurate, fine‑grained perception while preserving its original capabilities.
AI researchPlug and PlayVLM
0 likes · 15 min read

CodeGenie, HarmonyOS's AI coding assistant, now generates compilable ArkTS UI code from screenshots in ~2 minutes using a parse-then-generate architecture with VLM element detection and LLM agents, achieving 85%+ parsing accuracy, near 100% preview rate, and 40% adoption across food, travel, shopping, news, and education apps.
AI-assisted codingArkTSCodeGenie
0 likes · 14 min read

This article presents a step‑by‑step agentic object detection pipeline that combines open‑vocabulary detectors such as Grounding‑DINO with visual language models (GPT‑4o, o1) for concept extraction, critique, refinement, and validation, complete with code snippets, design rationale, and real‑world examples.
Grounding DINOOpen-Vocabulary DetectionPipeline
0 likes · 33 min read

The talk outlines Multimodal RAG implementation routes, comparing OCR‑based object recognition, transformer encoder‑decoder encoding, and Visual Language Model processing, explains the ColPali late‑interaction method for multi‑dimensional vector matching, addresses scaling tensors with binarization and reranking, and recommends a hybrid long‑term strategy where VLM excels on abstract imagery while traditional OCR remains valuable.
ColPaliDocument ProcessingMultimodal RAG
0 likes · 10 min read

This guide explains when to rely on prompt engineering versus SFT fine‑tuning for Vision‑Language Models, emphasizing data quality, appropriate dataset sizes, training epochs, hyper‑parameter tuning, and practical steps to build robust VLM pipelines.
AIDPOPrompt engineering
0 likes · 10 min read

This guide walks through setting up the open‑source Kotaemon framework, configuring Azure Document Intelligence and a visual large model, and implementing code to extract and caption images and tables from PDFs for end‑to‑end multimodal RAG applications.
AzurePythonRAG
0 likes · 12 min read
