# VWU.AI > Custom synthetic human-behavior video training data for AI models, with scenario design, video production, quality review, annotation and dataset delivery. VWU.AI is a project-based training-data service. The website describes custom data production rather than a fixed off-the-shelf dataset catalog. Deliverables and acceptance criteria are agreed per project. Homepage illustrations and sample videos demonstrate scenarios; they are not real-time monitoring or benchmark results. ## Site and services - [English homepage](https://vwu.ai/): Overview of custom synthetic video data, applications, production workflow, sample videos and contact options. - [Chinese homepage](https://vwu.ai/zh/): Chinese-language overview of the same services, workflow and samples. - [Driver and occupant monitoring data](https://vwu.ai/driver-monitoring-data.html): In-cabin scenarios including fatigue, distraction, phone use and occupant interactions, with sample-first production and labeling. - [驾驶员与乘员监控训练数据](https://vwu.ai/zh/driver-monitoring-data.html): Chinese service details covering in-cabin scenario variables, review, annotations and delivery. ## Plain-text reference - [Business and service overview](https://vwu.ai/llms-full.txt): A concise factual overview of the business, project workflow, data types and contact information. ## Optional - [XML sitemap](https://vwu.ai/sitemap.xml): Canonical public HTML pages and language relationships. ## Guides - [DMS and OMS training data procurement and acceptance checklist](https://vwu.ai/guides/dms-oms-training-data-checklist.html): Practical guidance on behavior definitions, negative examples, metadata, grouped evaluation splits and sample acceptance. - [驾驶员与乘员监控训练数据采购验收清单](https://vwu.ai/guides/zh/dms-oms-training-data-checklist.html): Chinese guide to behavior labels, timing boundaries, provenance, evaluation splits and acceptance. ## Service and procurement resources - [AI Video Annotation Services for Model Training](https://vwu.ai/ai-video-annotation.html) - [Custom AI Training Data Production Services](https://vwu.ai/custom-ai-training-data.html) - [AI Training Data Examples & Sample Evaluation](https://vwu.ai/datasets-and-case-studies.html) - [Human Action Recognition Video Data](https://vwu.ai/human-action-data.html) - [Driver Distraction Datasets: Labels and Hard Negatives](https://vwu.ai/resources/driver-distraction-dataset-annotation.html) - [Driver Drowsiness Datasets: Define Labels Before You Buy](https://vwu.ai/resources/driver-drowsiness-dataset-guide.html) - [Synthetic vs Real In-Cabin Data: Design a Useful Evaluation](https://vwu.ai/resources/synthetic-vs-real-in-cabin-data.html) - [Video Annotation Quality: A Practical Acceptance Checklist](https://vwu.ai/resources/video-annotation-quality-checklist.html) - [训练数据采购与验收指南](https://vwu.ai/resources/zh/) - [Training Data Procurement Guides](https://vwu.ai/resources.html) - [AI 视频标注服务与训练数据质检](https://vwu.ai/services/zh/ai-video-annotation.html) - [定制 AI 训练数据生产服务](https://vwu.ai/services/zh/custom-ai-training-data.html) - [AI 训练数据示例与样本评估](https://vwu.ai/services/zh/datasets-and-case-studies.html) - [人体动作识别视频训练数据](https://vwu.ai/services/zh/human-action-data.html) - [定制合成视频训练数据](https://vwu.ai/services/zh/synthetic-video-data.html) - [Custom Synthetic Video Training Data](https://vwu.ai/synthetic-video-data.html) Published sample media are illustrative synthetic scenarios. Field examples are illustrative specifications, not customer deliveries. Service feasibility, annotation scope and acceptance checks are agreed per project. ## Illustrated dataset split audit — 2026-10-29 - [Driver Monitoring Dataset Splits: Prevent Data Leakage](https://vwu.ai/resources/driver-monitoring-data-leakage-splits.html) - [驾驶员监控数据如何防止训练与测试集泄漏](https://vwu.ai/resources/zh/driver-monitoring-data-leakage-splits.html) Original workflow diagrams and illustrative audit code; no customer result or measured performance claim. ## Illustrated temporal annotation guide — 2026-10-31 - [In-Cabin Action Annotation: Define Temporal Boundaries](https://vwu.ai/resources/in-cabin-temporal-annotation-boundaries.html) - [车内行为时间标注:起止边界、遮挡与争议如何验收](https://vwu.ai/resources/zh/in-cabin-temporal-annotation-boundaries.html) Original boundary-labeling diagrams and a worked temporal IoU example; no customer result or measured performance claim. ## Illustrated multi-view synchronization guide — 2026-11-01 - [Multi-Camera In-Cabin Video: Verify Time Synchronization](https://vwu.ai/resources/multi-camera-in-cabin-video-synchronization.html) - [车内多视角视频同步:时间戳、偏移与漂移如何验收](https://vwu.ai/resources/zh/multi-camera-in-cabin-video-synchronization.html) Original synchronization workflow diagrams and illustrative timestamp arithmetic; no customer result or measured performance claim. ## Chinese procurement guides · 2026-11-02 - [驾驶员疲劳检测数据:采购前先定义标签](https://vwu.ai/resources/zh/driver-drowsiness-dataset-guide.html) - [驾驶员分心检测数据:行为标签与易混淆负例](https://vwu.ai/resources/zh/driver-distraction-dataset-annotation.html) - [合成与真实车内数据:如何设计有效评估](https://vwu.ai/resources/zh/synthetic-vs-real-in-cabin-data.html) - [视频标注质量:可复核的交付验收清单](https://vwu.ai/resources/zh/video-annotation-quality-checklist.html) ## Illustrated DMS false-alert evaluation guide — 2026-11-02 - [DMS False Alarms per Hour: An Event-Based Evaluation Guide](https://vwu.ai/resources/dms-false-alarms-per-hour-evaluation.html) - [驾驶员监控误报警评估:每小时次数与事件级验收](https://vwu.ai/resources/zh/dms-false-alarms-per-hour-evaluation.html) Original event-evaluation diagrams and fictional arithmetic; no customer result or measured performance claim. ## Illustrated keypoint visibility and export acceptance guide — 2026-11-04 - [In-Cabin Keypoints: Visibility and Export Acceptance](https://vwu.ai/resources/in-cabin-keypoint-visibility-export-checks.html) - [车内关键点数据验收:可见性状态与导出回读](https://vwu.ai/resources/zh/in-cabin-keypoint-visibility-export-checks.html) Original state and roundtrip diagrams; proposed acceptance protocol, not customer results or measured performance. - [多模态视觉语言模型的夜间车内数据:红外样本怎么验收](https://vwu.ai/resources/zh/multimodal-vlm-night-infrared-data-evaluation.html) - [Night and Infrared Cabin Data for Multimodal VLM Evaluation](https://vwu.ai/resources/multimodal-vlm-night-infrared-data-evaluation.html) - [视觉语言模型的驾驶员视线区域标注:让回答对应车内位置](https://vwu.ai/resources/zh/vlm-driver-gaze-zone-grounding.html) - [Grounding Driver Gaze Zones for Vision-Language Models](https://vwu.ai/resources/vlm-driver-gaze-zone-grounding.html) - [视频理解与视觉语言模型:手机使用识别的困难负样本怎么设计](https://vwu.ai/resources/zh/video-vlm-phone-use-hard-negatives.html) - [Hard Negatives for Phone-Use Recognition in Video VLMs](https://vwu.ai/resources/video-vlm-phone-use-hard-negatives.html) - [具身智能与车内视频合成数据:如何建立长尾场景覆盖矩阵](https://vwu.ai/resources/zh/synthetic-data-long-tail-coverage-matrix.html) - [Synthetic Data: Long-Tail Coverage for Embodied AI](https://vwu.ai/resources/synthetic-data-long-tail-coverage-matrix.html) - [多模态视频训练数据交付:清单、哈希与来源记录怎么验收](https://vwu.ai/resources/zh/multimodal-video-dataset-manifest-integrity.html) - [Multimodal Video Dataset Delivery: Manifests, Checksums and Provenance](https://vwu.ai/resources/multimodal-video-dataset-manifest-integrity.html) - [Cosmos 更新后,世界模型训练数据该怎样组织与验收](https://vwu.ai/resources/zh/cosmos-world-model-data-recipe.html) ## Driver monitoring dataset cost - [Driver Monitoring Dataset Cost: 8 Specifications to Define Before a Quote](https://vwu.ai/resources/driver-monitoring-dataset-cost.html) - [车内监控训练数据多少钱?报价前需要明确的 8 项规格](https://vwu.ai/resources/zh/driver-monitoring-dataset-cost.html) - [車内モニタリング学習データの費用:見積もり前に決める8つの仕様](https://vwu.ai/resources/ja/driver-monitoring-dataset-cost.html) - [Waymo 冬季准备带来的启发:视觉数据集如何写清适用边界](https://vwu.ai/resources/zh/winter-driving-visual-data-boundaries.html) - [从 Isekai3D 看配对视觉数据:多视角、深度与动作标签如何对齐](https://vwu.ai/resources/zh/isekai3d-paired-visual-data.html) - [韩国 Physical AI 新进展:机器人视频数据如何形成可追溯闭环](https://vwu.ai/resources/zh/physical-ai-robot-data-loop.html) - [AI Agent 开始搭建仿真场景,生成视频怎样验收才可靠](https://vwu.ai/resources/zh/agent-generated-simulation-acceptance.html) - [Toyota 与 Waymo 扩展合作,车内视觉数据跨市场迁移该检查什么](https://vwu.ai/resources/zh/cross-market-cabin-data-validation.html) - [城市视频 AI 应用新进展:事件数据集怎样保留证据与隐私边界](https://vwu.ai/resources/zh/video-ai-event-evidence-privacy.html) - [公开 DMS 数据还是定制数据?采购决策指南](https://vwu.ai/resources/zh/public-vs-custom-driver-monitoring-data.html) - [Public or Custom DMS Data? A Procurement Decision Guide](https://vwu.ai/resources/public-vs-custom-driver-monitoring-data.html) - [公開DMSデータか個別制作か:調達判断ガイド](https://vwu.ai/resources/ja/public-vs-custom-driver-monitoring-data.html) - [DMS 数据试单怎么验收?批量生产前的计划表](https://vwu.ai/resources/zh/dms-data-pilot-acceptance-plan.html) - [DMS Data Pilot Acceptance: A Plan Before Batch Production](https://vwu.ai/resources/dms-data-pilot-acceptance-plan.html) - [DMSデータの試作検収:量産前に決める確認計画](https://vwu.ai/resources/ja/dms-data-pilot-acceptance-plan.html) - [车内相机位置变了,训练数据需要重做吗?](https://vwu.ai/resources/zh/driver-monitoring-camera-placement-data.html) - [DMS Camera Placement Changes: A Training-Data Review Plan](https://vwu.ai/resources/driver-monitoring-camera-placement-data.html) - [DMSカメラの位置変更:学習データを見直す手順](https://vwu.ai/resources/ja/driver-monitoring-camera-placement-data.html) - [合成视频看着真实就能训练吗?时序一致性检查指南](https://vwu.ai/resources/zh/synthetic-video-temporal-consistency-review.html) - [Synthetic Video Quality: A Temporal Consistency Review](https://vwu.ai/resources/synthetic-video-temporal-consistency-review.html) - [合成動画の品質:時間的な一貫性を確認する方法](https://vwu.ai/resources/ja/synthetic-video-temporal-consistency-review.html)