Results for “vision-transformer”

Load pre-trained models from Hugging Face Hub, run pipeline inference, generate text, and fine-tune models on NLP, vision, audio, and multimodal tasks using the Transformers library.

Import object detection models from HuggingFace or NVIDIA NGC into a DeepStream pipeline with automated ONNX download, TensorRT engine build, custom parser, multi-stream benchmark, and PDF report generation.

Trains, evaluates, and runs inference for BEVFusion multi-sensor 3D object detection models that fuse LiDAR and camera data in bird's-eye-view space for autonomous driving.

Deploy and operate the RTVI-CV-3D microservice for multi-camera 3D detection and tracking, supporting sample datasets, custom videos, and RTSP streams.

Computer vision engineering skill for object detection, image segmentation, and visual AI systems. Covers CNN and Vision Transformer architectures, YOLO/Faster R-CNN/DETR detection, Mask R-CNN/SAM segmentation, and production deployment with ONNX/TensorRT. Includes PyTorch, torchvision, Ultralytics, Detectron2, and MMDetection frameworks. Use when building detection pipelines, training custom models, optimizing inference, or deploying vision systems.

Generate image captions, answer visual questions, and perform image-text retrieval using BLIP-2's Q-Former architecture with frozen vision encoders and LLMs.

Enables visual instruction tuning and image-based conversations using open-source vision-language models. Supports multi-turn image chat, visual question answering, and image understanding tasks.

Evaluates text-to-image models on spatial relationship accuracy using the VISOR metric, separating object detection from spatial correctness to reveal biases like object priority and merging.

VS-Enhanced Theoretical Framework Architect with Critique & Visualization Full VS 5-Phase process: Modal theory avoidance, Long-tail exploration, differentiated framework presentation Absorbed A3 (Devil's Advocate) critique and A6 (Conceptual Framework Visualizer) capabilities Use when: building theoretical foundations, designing conceptual models, deriving hypotheses, critiquing frameworks, visualizing models Triggers: theoretical framework, 이론적 프레임워크, conceptual model, 개념적 모형, hypothesis derivation, critique, devil's advocate, 반론, visualization, diagram

Donut: Document Understanding Transformer without OCR

Visual Prompt Tuning

Trains, evaluates, exports, and runs inference for Visual ChangeNet models used in AOI defect detection, comparing image pairs for PASS/NO_PASS classification or change-segmentation masks.

Train, evaluate, distill, quantize, export, and run inference for RT-DETR object detection models using NVIDIA TAO.

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP vision encoder with Vicuna/LLaMA language models. Supports multi-turn image chat, visual question answering, and instruction following. Use for vision-language chatbots or image understanding tasks. Best for conversational image analysis.

Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints and activation caching. Use when reverse-engineering model algorithms, studying attention patterns, or performing activation patching experiments.

Pixtral 12B: A Frontier Multimodal Model

Runs the open-source LLaVA vision-language model for image understanding, captioning, visual question answering, and multi-turn image conversations, including setup, inference, and training guidance.

Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP vision encoder with Vicuna/LLaMA language models. Supports multi-turn image chat, visual question answering, and instruction following. Use for vision-language chatbots or image understanding tasks. Best for conversational image analysis.

Inspect and manipulate transformer internals via HookPoints and activation caching for mechanistic interpretability research.

Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints and activation caching. Use when reverse-engineering model algorithms, studying attention patterns, or performing activation patching experiments.

Extend context windows of transformer models using RoPE, YaRN, ALiBi, and position interpolation techniques. Use when processing long documents (32k-128k+ tokens), extending pre-trained models beyond original context limits, or implementing efficient positional encodings. Covers rotary embeddings, attention biases, interpolation methods, and extrapolation strategies for LLMs.

画像を理解（被写体・テキストOCR・構図・色・UI構造の分析）し、結果を構造化して返すスキル。CC CLI は GLM-5.3 等の vision 非対応モデルで稼働中のため画像を直接視認できず、主ルート Gemini 2.5 Flash（scripts/api/gemini_vision.py・無料枠）と副ルート 4_5v MCP（analyze_image・Readが返すCDN URL）の2経路で分析し、CCは結果の構造化・比較・保存に専任する。 ユーザーが「画像見て」「この画像何が写ってる」「画像比較して」「スクショ見て」「画像分析して」「画像理解」「vision-analyze」と言った時、または /vision-analyze を呼んだ時にトリガー。 ※画像生成（image generation）は対象外（make-song / video-prompt-spec / demo-site-sales参照）。ピクセル修正（花鈿除去等）は remove-huadian の役割。楽曲分析は analyze-song / reverse-engineer-song。

v5.6.0 joint multi-TF model: single model per symbol with broadcast 1Hour context replaces dual 15Min/1Hour models. Trigger: (1) replacing weighted-voting model aggregation, (2) adding broadcast features to vectorized env, (3) limited training data + worried about overfitting from doubling obs_dim, (4) backtest builder mismatch with newer feature counts.