Computer Vision
Image recognition, video understanding, multimodal AI
221 articles across 72 editions
Articles
- Qwen 3.5 vs Gemini 3 Pro on Screenshot-to-Code: Is the gap finally gone? -- 2026-02-19
- [Editorial] The evolution of vision models from CNNs to transformers -- 2026-02-19
- Show HN: Offline tiles and routing and geocoding in one Docker Compose stack -- 2026-01-05
- ostris/Z-Image-De-Turbo -- 2025-12-12
- mistralai/Ministral-3-8B-Instruct-2512 -- 2025-12-12
- Making Glasses That Detect Smartglasses -- 2025-12-11
- ByteDance-Seed/Depth-Anything-3 -- 2025-12-10
- SARLO-80: Worldwide Slant SAR Language Optic Dataset at 80 cm Resolution -- 2025-12-10
- Optical Context Compression Is Just (Bad) Autoencoding -- 2025-12-10
- seominseok0429/Upsample-Anything-A-Simple-and-Hard-to-Beat-Baseline-for-Feature-Upsampling -- 2025-12-09
- Vision-and-Language Navigation with Analogical Textual Descriptions in LLMs -- 2025-12-09
- lrzjason/QwenEdit-Anything2Real_Alpha -- 2025-12-08
- How Big is Your Video Again? Square vs Rectangular Pixels -- 2025-12-08
- ByteDance/BindWeave -- 2025-12-08
- 3D Gaussian and Diffusion-Based Gaze Redirection -- 2025-12-04
- apple/starflow -- 2025-12-04
- princepainter/ComfyUI-PainterLongVideo -- 2025-12-04
- OpenGPT-4o-Image: A Comprehensive Dataset for Advanced Image Generation and Editing -- 2025-12-03
- Introducing GeoVista: Web-Augmented Agentic Visual Reasoning for Geolocalization | "GeoVista is a new 7B open-source agentic model that achieves SOTA performance in geolocalization by integrating visual tools and web search into an RL loop." -- 2025-11-28
- facebook/sam-3d-body-dinov3 -- 2025-11-28
- Diffusers welcomes FLUX-2 -- 2025-11-26
- marinero4972/Open-o3-Video -- 2025-11-17
- nanonets/Nanonets-OCR2-3B -- 2025-11-17
- nvidia/ChronoEdit-14B-Diffusers-Upscaler-Lora -- 2025-11-17
- We ran over 600 image generations to compare AI image models -- 2025-11-13
- dx8152/Qwen-Image-Edit-2509-Relight -- 2025-11-13
- Last week in Multimodal AI - Local Edition -- 2025-11-12
- DeepSeek-OCR GGUF model runs great locally - simple and fast -- 2025-11-12
- Qwen3-VL works really good with Zoom-in Tool -- 2025-11-12
- lightonai/LightOnOCR-1B-1025 -- 2025-11-12
- Qwen/Qwen3-VL-2B-Thinking -- 2025-11-12
- Why does Image Recognition work in llama-server but not through Open WebUI? -- 2025-11-06
- Has anyone tested ollama on Whisplay HAT with Raspberry pi zero 2W? -- 2025-11-06
- allenai/olmOCR-2-7B-1025 -- 2025-11-06
- is there simple way like .bat to compress to q4-q8 like Unsloth, Qwen3-VL-30B-A3B-Thinking-abliterated model -- 2025-11-06
- meituan-longcat/LongCat-Video -- 2025-11-05
- allenai/olmOCR-2-7B-1025-FP8 -- 2025-11-05
- Worse Embedding Performance with Qwen 3 VL than with Qwen 2.5 VL? -- 2025-11-05
- Retrospective Sparse Attention for Efficient Long-Context Generation -- 2025-11-05
- deepseek-ai/DeepSeek-OCR -- 2025-11-04
- LiquidAI/LFM2-VL-3B -- 2025-11-04
- Qwen/Qwen3-VL-235B-A22B-Thinking -- 2025-11-04
- KTransformers Open Source New Era: Local Fine-tuning of Kimi K2 and DeepSeek V3 -- 2025-11-04
- F2LLM Technical Report: Matching SOTA Embedding Performance with 6 Million Open-Source Data -- 2025-11-04
- [Editorial] https://blog.peerllm.com/2025/11/02/announcing-v0.7.6.html -- 2025-11-04
- Myths Programmers Believe about CPU Caches -- 2025-11-04
- Pi Zero Powers A Little Indoor Rover -- 2025-11-04
- [Editorial] https://commsrisk.com/sms-blaster-and-imsi-catcher-news-from-lebanon-cambodia-switzerland-and-the-philippines/ -- 2025-11-03
- An Obscure Military Program Helps Local Cops Buy Armored Card and Spyware -- 2025-11-03
- DeepSeek may have found a new way to improve AI’s ability to remember -- 2025-11-02
- Qwen/Qwen3-VL-8B-Thinking -- 2025-11-02
- nvidia/omnivinci -- 2025-11-02
- OpenImagingLab/FlashVSR -- 2025-11-02
- Build Your Own Force-Feedback Joystick -- 2025-11-02
- DarkBitx/ICRev -- 2025-11-01
- dd1100/DiscordRAT -- 2025-11-01
- Police used Flock cameras to accuse a woman of theft, she had to prove innocence -- 2025-11-01
- ZOZO's Contact Solver for physics-based simulations -- 2025-11-01
- valiantcat/Qwen-Image-Edit-MeiTu -- 2025-11-01
- ToonComposer: Streamlining Cartoon Production with Generative Post-Keyframing -- 2025-11-01
- DeepSeek just released a bombshell AI model (DeepSeek AI) so profound it may be as important as the initial release of ChatGPT-3.5/4 ------ Robots can see-------- And nobody is talking about it -- And it's Open Source - If you take this new OCR Compresion + Graphicacy = Dual-Graphicacy 2.5x improve -- 2025-10-27
- Pico Banana: Large-Scale Dataset for Image Editing by Apple -- 2025-10-27
- What's the best embedding model for document images ? -- 2025-10-26
- AlphaXiv,Compare the Deepseek-OCR and Mistral-OCR OCR models -- 2025-10-26
- Open-Bee/Bee-8B-RL -- 2025-10-26
- datalab-to/chandra -- 2025-10-26
- Unlock the power of images with AI Sheets -- 2025-10-26
- dvlab-research/DreamOmni2 -- 2025-10-25
- bytetriper/RAE -- 2025-10-25
- tencent/POINTS-Reader -- 2025-10-25
- Stitch: Training-Free Position Control in Multimodal Diffusion Transformers -- 2025-10-25
- mit-han-lab/streaming-vlm -- 2025-10-22
- Doby-Xu/WithAnyone -- 2025-10-22
- lightx2v/Wan2.2-I2V-A14B-Moe-Distill-Lightx2v -- 2025-10-22
- tencent-ailab/SongPrep -- 2025-10-20
- opendatalab/MinerU2.5-2509-1.2B -- 2025-10-20
- QuantStack/Qwen-Image-Edit-2509-GGUF -- 2025-10-20
- LM Studio and VL models -- 2025-10-19
- Qwen/Qwen-Image-Edit-2509 -- 2025-10-19
- Alpha-VLLM/Lumina-DiMOO -- 2025-10-19
- Paper2Video — turn a research paper into a full presentation video (slides, speech, talking head) -- 2025-10-15
- Practical OCR with Nanonets OCR2‑3B -- 2025-10-15
- neuphonic/neutts-air -- 2025-10-15
- Qwen/Qwen3-VL-235B-A22B-Instruct -- 2025-10-15
- XiaomiMiMo/MiMo-Audio-Eval -- 2025-10-15
- Very interesting! OmniInsert — mask-free video insertion of any reference -- 2025-10-14
- facebookresearch/DepthLM_Official -- 2025-10-14
- NVlabs/rcm -- 2025-10-11
- Divining Air Quality With A Cheap Computer Vision Device -- 2025-10-09
- pixai-labs/pixai-tagger-v0.9 -- 2025-10-09
- TianDongL/Diffusion_pipe_in_ComfyUI -- 2025-10-09
- Project running VLMs on a Pi 5 and NV Jetson Orin Nano -- 2025-10-05
- Demo: I made an open-source version of Imagine by Claude (released yesterday) -- 2025-10-05
- nunchaku-tech/nunchaku-qwen-image-edit-2509 -- 2025-10-05
- cvlab-kaist/VIRAL -- 2025-10-05
- deepseek-ai/DeepSeek-V3.2-Exp -- 2025-10-03
- moondream/moondream3-preview -- 2025-10-03
- Unsupervised Hallucination Detection by Inspecting Reasoning Processes -- 2025-10-03
- Learning to See Before Seeing: Demystifying LLM Visual Priors from Language Pre-training -- 2025-10-03
- Will fine-tuning LLaMA 3.2 11B Instruct on text-only data degrade its vision capabilities? -- 2025-10-03
- jmanhype/vggt-mps -- 2025-10-02
- openbmb/VoxCPM-0.5B -- 2025-10-02
- Comfy-Org/Qwen-Image-Edit_ComfyUI -- 2025-10-02
- SOTA OCR on-device with Core ML and dots.ocr -- 2025-10-02
- Stress-Testing RAG in Production: Retrieval Quality, Drift, and Hidden Costs -- 2025-09-30
- Tencent-Hunyuan/SRPO -- 2025-09-30
- lodestones/Chroma1-Base -- 2025-09-30
- MV-RAG: Retrieval Augmented Multiview Diffusion -- 2025-09-30
- Phantom-video/HuMo -- 2025-09-27
- Build Your Own 6K Camera -- 2025-09-27
- Training-Free Text-Guided Color Editing with Multi-Modal Diffusion Transformer -- 2025-09-27
- OPPOer/Qwen-Image-Pruning -- 2025-09-27
- From Research to Reality: Feasibility of Gradient Inversion Attacks in Federated Learning -- 2025-09-17
- Renting GPUs is hilariously cheap -- 2025-09-09
- Tencent-Hunyuan/HunyuanWorld-Voyager -- 2025-09-09
- Shipping textures as PNGs is suboptimal -- 2025-09-09
- inclusionAI/UI-Venus -- 2025-09-07
- WeChatCV/Stand-In_Preprocessor_ComfyUI -- 2025-09-07
- Robotic Canoe Puts Robot Arms to Work -- 2025-09-07
- LiquidAI/LFM2-VL-1.6B -- 2025-09-06
- A Training-Free, Task-Agnostic Framework for Enhancing MLLM Performance on High-Resolution Images -- 2025-09-06
- TencentARC/GenCompositor -- 2025-09-06
- Doubling Your Data in Minutes: Ultra-fast Tabular Data Generation via LLM-Induced Dependency Graphs -- 2025-08-30
- Temporal Point-Supervised Signal Reconstruction: A Human-Annotation-Free Framework for Weak Moving Target Detection -- 2025-08-30
- PurinNyova/Image-Detection-Bypass-Utility -- 2025-08-25
- An Alternative to Text-to-SQL -- 2025-08-25
- Best model for transcribing videos? -- 2025-08-25
- Meta released DINO-V3 : SOTA for any Vision task -- 2025-08-21
- DatasetAgent: A Novel Multi-Agent System for Auto-Constructing Datasets from Real-World Images -- 2025-08-21
- WeChatCV/Stand-In -- 2025-08-21
- We built a 12B model that beats Claude 4 Sonnet at video captioning while costing 17x less - fully open source -- 2025-08-15
- Francis-Rings/StableAvatar -- 2025-08-15
- nvidia/canary-qwen-2.5b -- 2025-08-15
- OmniSVG/OmniSVG -- 2025-08-15
- Phi-Ground Tech Report: Advancing Perception in GUI Grounding -- 2025-08-15
- [UPDATE] DocStrange - Structured data extraction from images/pdfs/docs -- 2025-08-15
- NuMarkdown-8B-Thinking - first reasoning OCR VLM -- 2025-08-11
- Geometry Forcing: Marrying Video Diffusion and 3D Representation for Consistent World Modeling -- 2025-08-11
- Vision Language Model Alignment in TRL ⚡️ -- 2025-08-11
- How the best image generation models work from the inside ? -- 2025-08-08
- AIDC-AI/Ovis-U1-3B -- 2025-08-08
- n0xa/SecKC-MHN-Globe -- 2025-08-08
- LLM-Based Identification of Infostealer Infection Vectors from Screenshots: The Case of Aurora -- 2025-08-08
- Reason ex Machina: Jailbreaking LLMs by Squeezing Their Brains | xayan.nu -- 2025-08-08
- Closing the Modality Gap for Mixed Modality Search -- 2025-08-07
- [Editorial] Mach-O binary analysis, with a focus on malware analysis and reverse engineering. -- 2025-07-31
- petqoo/ROGO -- 2025-07-31
- wyhlovecpp/GPT-Image-Edit -- 2025-07-31
- Show HN: MoebiusXBIN – ASCII and text-mode art editor with custom font support -- 2025-07-31
- boson-ai/higgs-audio -- 2025-07-24
- nvidia/canary-qwen-2.5b -- 2025-07-24
- TimeScope: How Long Can Your Video Large Multimodal Model Go? -- 2025-07-24
- ICML 2025 Outstanding Paper Awards -- 2025-07-24
- THUDM/GLM-4.1V-Thinking -- 2025-07-23
- FunAudioLLM/ThinkSound -- 2025-07-23
- RaphaelLiu/PusaV1 -- 2025-07-23
- merve/smol-vision -- 2025-07-23
- ChenDarYen/ComfyUI-NAG -- 2025-07-20
- Locality-aware Parallel Decoding for Efficient Autoregressive Image Generation -- 2025-07-20
- quasiblob/ComfyUI-EsesImageEffectBloom -- 2025-07-18
- HiDream-ai/HiDream-E1-1 -- 2025-07-18
- Autoregressive Image Generation with Linear Complexity: A Spatial-Aware Decay Perspective -- 2025-07-18
- runjiali-rl/vmem -- 2025-07-17
- GeoArrow and GeoParquet, and the Future of Geospatial Data Analysis -- 2025-07-15
- TextPixs: Glyph-Conditioned Diffusion with Character-Aware Attention and OCR-Guided Supervision -- 2025-07-15
- Need advice on how to improve Handwritten Text Recognition of names using Vision models (for academic research purposes) -- 2025-07-14
- A fast 3D collision detection algorithm -- 2025-07-11
- 1000 African Voices: Advancing inclusive multi-speaker multi-accent speech synthesis -- 2025-07-06
- Hack Swaps Keys for Gang Signs, Everyone Gets In -- 2025-07-06
- Marker Gene Method : Identifying Stable Solutions in a Dynamic Environment -- 2025-07-06
- bytedance/ATI -- 2025-07-01
- AIDC-AI/Ovis-U1-3B -- 2025-07-01
- google/gemma-3n-E4B-it -- 2025-07-01
- baidu/ERNIE-4.5-21B-A3B-PT -- 2025-07-01
- bullerwins/FLUX.1-Kontext-dev-GGUF -- 2025-06-28
- google/gemma-3n-E2B-it -- 2025-06-28
- unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF -- 2025-06-26
- jinaai/jina-embeddings-v4 -- 2025-06-26
- Intelligent-Internet/II-Medical-8B-1706 -- 2025-06-26
- Tencent-Hunyuan/HunyuanPortrait -- 2025-06-24
- (0,2) hybrid models -- 2025-06-24
- 0-th Order Pseudo-differential Operator on the Circle -- 2025-06-24
- OmniGen2/OmniGen2 -- 2025-06-24
- gdhe17/Self-Forcing -- 2025-06-24
- lightx2v/Wan2.1-T2V-14B-StepDistill-CfgDistill -- 2025-06-20
- Kijai/WanVideo_comfy -- 2025-06-20
- tencent/Hunyuan3D-2.1 -- 2025-06-20
- llama-server is cooking! gemma3 27b, 100K context, vision on one 24GB GPU. -- 2025-06-18
- I built a lightweight, private, MCP server to share context between AI tools -- 2025-06-18
- KVzip: Query-agnostic KV Cache Eviction — 3~4× memory reduction and 2× lower decoding latency -- 2025-06-18
- Ollama now supports streaming responses with tool calling -- 2025-06-18
- Chainlit or Open webui for production? -- 2025-06-18
- Ollama not releasing VRAM after running a model -- 2025-06-18
- Help Shape the Future of AI in India - Survey on Local vs Cloud LLM Usage (Developers/Students/AI Enthusiasts) -- 2025-06-18
- llmcontext: Attach you whole project in large context chats -- 2025-06-18
- Semantic search engine for ArXiv, biorxiv and medrxiv -- 2025-06-18
- Oodle 2.9.14 and Intel 13th/14th gen CPUs -- 2025-06-18
- A multi-turn tool-calling base model for RL agent training -- 2025-06-18
- GUI RAG that can do an unlimited number of documents, or at least many -- 2025-06-18
- showlab/OmniConsistency -- 2025-06-15
- echo840/MonkeyOCR -- 2025-06-15
- inclusionAI/Ming-Lite-Omni -- 2025-06-15
- New method for creating large 3D models of urban areas is faster and cheaper -- 2025-06-15
- showlab/D-AR -- 2025-06-14
- graphdeco-inria/on-the-fly-nvs -- 2025-06-14
- YOLO-World: Real-Time Open-Vocabulary Object Detection -- 2025-06-13
- 0ptical trapping with optical magnetic field and photonic Hall effect forces -- 2025-06-13
- (0,2) Mirror Symmetry on homogeneous Hopf surfaces -- 2025-06-13
- 0/1 Deep Neural Networks via Block Coordinate Descent -- 2025-06-10
- Hcompany/Holo1-3B -- 2025-06-04
- black-forest-labs/FLUX.1-schnell -- 2025-06-04
- PlayHT/PlayDiffusion -- 2025-06-04
- AMAP-ML/UniVG-R1 -- 2025-06-02
- showlab/OmniConsistency -- 2025-06-02
- tencent/HunyuanVideo-Avatar -- 2025-06-02
- Datadog/Toto-Open-Base-1.0 -- 2025-06-01
- 0.08 fF, 0.72 nA dark current, 91% Quantum Efficiency, 38 Gb/s Nano-photodetector on a 45 nm CMOS Silicon-Photonic Platform -- 2025-05-30
- 1000 FPS HDR Video With a Spike-RGB Hybrid Camera -- 2025-05-30
- 100,000 frames-per-second compressive imaging with a conventional rolling-shutter camera by random point-spread-function engineering -- 2025-05-29
- 1,000-Fold Enhancement of Light-Induced Magnetism in Plasmonic Au Nanoparticles -- 2025-05-29
- Model-Based Machine Learning (2023) -- 2025-05-28
- 0-MMS: Zero-Shot Multi-Motion Segmentation With A Monocular Event Camera -- 2025-05-28
- 0.8% Nyquist computational ghost imaging via non-experimental deep learning -- 2025-05-28