在低風(fēng)險(xiǎn)認(rèn)知狀態(tài)下，駕駛員的注意力主要集中在前方區(qū)域和后視鏡上，注視和掃視頻率較低，注視點(diǎn)在各區(qū)域間循環(huán)轉(zhuǎn)移，掃視路徑呈現(xiàn)出明顯的規(guī)律性，如圖11所示。

圖11 眼動(dòng)軌跡與熱點(diǎn)圖（低風(fēng)險(xiǎn)認(rèn)知狀態(tài)）

在高風(fēng)險(xiǎn)認(rèn)知狀態(tài)下，駕駛員往往傾向于將注意力集中在與自車交互的周圍車輛上。對應(yīng)有更高的注視和掃視頻率，并呈現(xiàn)出重復(fù)注視和來回掃視的特點(diǎn)，如圖12所示。

圖12 眼動(dòng)軌跡與熱點(diǎn)圖（高風(fēng)險(xiǎn)認(rèn)知狀態(tài)）

5、總結(jié)與展望

針對駕駛數(shù)據(jù)集缺少與人類認(rèn)知相關(guān)信息的問題，研究采用駕駛模擬器實(shí)驗(yàn)的方法構(gòu)建RISEE數(shù)據(jù)集，提供自然駕駛場景、人類主觀風(fēng)險(xiǎn)認(rèn)知及眼動(dòng)等數(shù)據(jù)。

對場景主客觀行駛風(fēng)險(xiǎn)及眼動(dòng)特征的分析表明，RISEE數(shù)據(jù)集涵蓋各種交互類型的數(shù)據(jù)，在決策規(guī)劃系統(tǒng)的研發(fā)和測試中具有潛在的應(yīng)用價(jià)值。具體應(yīng)用包括：

基于風(fēng)險(xiǎn)認(rèn)知的決策規(guī)劃：行駛過程中，駕駛員通過視覺感知周圍環(huán)境，并在評估各類環(huán)境信息后形成認(rèn)知風(fēng)險(xiǎn)。因此，眼動(dòng)數(shù)據(jù)可用于識別影響人類風(fēng)險(xiǎn)認(rèn)知的關(guān)鍵環(huán)境因素，揭示風(fēng)險(xiǎn)認(rèn)知的內(nèi)在規(guī)律，從而幫助決策規(guī)劃系統(tǒng)發(fā)出與人類認(rèn)知相協(xié)調(diào)的決策指令。智能駕駛舒適性評估：在高等級智能駕駛汽車中，駕駛員角色發(fā)生了變化，對舒適性的評價(jià)也不再局限于車輛本身，而是會受到認(rèn)知風(fēng)險(xiǎn)的影響[1]。因此，可以基于RISEE探究眼動(dòng)信息與駕駛員主觀認(rèn)知風(fēng)險(xiǎn)之間的關(guān)系，評估高等級智能駕駛的舒適性。

參考文獻(xiàn)：

[1] Meng H, Zhao X, Chen J, et al. Study on physiological representation of passenger cognitive comfort: An example with overtaking scenarios[J]. Transportation research part F: traffic psychology and behaviour, 2024, 102: 241-259.

[2] Bao N, Carballo A, Tsukada M, et al. Personalized causal factor generalization for subjective risky scene understanding with vision transformer[C]//2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC). IEEE, 2023: 4637-4643.

[3] You S, Luo X, Liang X, et al. A Comprehensive LLM-powered framework for Driving Intelligence evaluation[J]. arXiv preprint arXiv:2503.05164, 2025.

[4] He X, Stapel J, Wang M, et al. Modelling perceived risk and trust in driving automation reacting to merging and braking vehicles[J]. Transportation research part F: traffic psychology and behaviour, 2022, 86: 178-195.

[5] Ke Z, Jiang Y, Wang Y, et al. D2E: An Autonomous Decision-Making Dataset involving Driver States and Human evaluation of Driving Behavior[C]//2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC). IEEE, 2024: 2294-2301.

[6] Vlakveld W P. A comparative study of two desktop hazard perception tasks suitable for mass testing in which scores are not based on response latencies[J]. Transportation research part F: traffic psychology and behaviour, 2014, 22: 218-231.

[7] Charlton S G, Starkey N J, Perrone J A, et al. What’s the risk? A comparison of actual and perceived driving risk[J]. Transportation research part F: traffic psychology and Behaviour, 2014, 25: 50-64.

[8] Wu X, Xing X, Chen J, et al. Risk assessment method for driving scenarios of autonomous vehicles based on drivable area[C]//2022 IEEE 25th international conference on intelligent transportation systems (ITSC). IEEE, 2022: 2206-2213.

責(zé)編丨高炳釗