ADAS-TO: A Large-Scale Multimodal Naturalistic Dataset and Empirical Characterization of Human Takeovers during ADAS Engagement
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910044889546752 |
|---|---|
| author | Wang, Yuhang Xu, Yiyao Sun, Jingran Zhou, Hao |
| author_facet | Wang, Yuhang Xu, Yiyao Sun, Jingran Zhou, Hao |
| contents | Takeovers remain a key safety vulnerability in production ADAS, yet existing public resources rarely provide takeover-centered, real-world data. We present ADAS-TO, the first large-scale naturalistic dataset dedicated to ADAS-to-manual transitions, containing 15,659 takeover-centered 20s clips from 327 drivers across 22 vehicle brands. Each clip synchronizes front-view video with CAN logs. Takeovers are defined as ADAS ON $\rightarrow$ OFF transitions, with the primary trigger labeled as brake, steer, gas, mixed, or system disengagement. We further separate planned driver-initiated terminations (Ego) from forced takeovers (Non-ego) using a rule-based partition. While most events occur within conservative kinematic margins, we identify a long tail of 285 safety-critical cases. For these events, we combine kinematic screening with vision--language (VLM) annotation to attribute hazards and relate them to intervention dynamics. The resulting cross-modal analysis shows distinct kinematic signatures across traffic dynamics, infrastructure degradation, and adverse environments, and finds that in 59.3% of critical cases, actionable visual cues emerge at least 3s before takeover, supporting the potential for semantics-aware early warning beyond late-stage kinematic triggers. The dataset is publicly released at huggingface.co/datasets/HenryYHW/ADAS-TO-Sample. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_06986 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ADAS-TO: A Large-Scale Multimodal Naturalistic Dataset and Empirical Characterization of Human Takeovers during ADAS Engagement Wang, Yuhang Xu, Yiyao Sun, Jingran Zhou, Hao Human-Computer Interaction Computer Vision and Pattern Recognition Takeovers remain a key safety vulnerability in production ADAS, yet existing public resources rarely provide takeover-centered, real-world data. We present ADAS-TO, the first large-scale naturalistic dataset dedicated to ADAS-to-manual transitions, containing 15,659 takeover-centered 20s clips from 327 drivers across 22 vehicle brands. Each clip synchronizes front-view video with CAN logs. Takeovers are defined as ADAS ON $\rightarrow$ OFF transitions, with the primary trigger labeled as brake, steer, gas, mixed, or system disengagement. We further separate planned driver-initiated terminations (Ego) from forced takeovers (Non-ego) using a rule-based partition. While most events occur within conservative kinematic margins, we identify a long tail of 285 safety-critical cases. For these events, we combine kinematic screening with vision--language (VLM) annotation to attribute hazards and relate them to intervention dynamics. The resulting cross-modal analysis shows distinct kinematic signatures across traffic dynamics, infrastructure degradation, and adverse environments, and finds that in 59.3% of critical cases, actionable visual cues emerge at least 3s before takeover, supporting the potential for semantics-aware early warning beyond late-stage kinematic triggers. The dataset is publicly released at huggingface.co/datasets/HenryYHW/ADAS-TO-Sample. |
| title | ADAS-TO: A Large-Scale Multimodal Naturalistic Dataset and Empirical Characterization of Human Takeovers during ADAS Engagement |
| topic | Human-Computer Interaction Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.06986 |