驭模有道，智胜未来：大模型助力智能化变革

F o u n d a t i o nm o d e l f a c i l i t a t e t h e i n t e l l i g e n t t r a n s f o r m a t i o n 陈勇博士Dr.Yong Chen吉利汽车GEELY AUTO 基于用户体验驱动技术价值创造，使智能化设计回归理性Createtechnological value driven byuser experience,bringing intelligent design back to rationality. 大模型发展核心四要素:3+1The four core elements of the development offoundation models: 3+1 人工智能时代：涌现式+继承式A r t i f i c i a lI n t e l l i g e n c eE r a：E m e r g e n t+I n h e r i t e d 算法Algorithm ◼LLM / M ultimodal M odel◼AI-DRIVE智能驾驶大模型◼AI-D RI VE Intelligent Driv ingFoundation M odel 大模型助力智能化变革Foundationmodels boost the intelligent transformation 丰富的生成内容Rich Generative Content 用户场景决定技术价值User scenarios determine the value of technology W ea l l n e e d f o u n d a t i o n m o d e ls D ow e a l l n e e d f o u n d a t i o n m o d e ls ? 产品需要有市场和价值定位The product needs to have a market and value positioning 新技术是来解决问题的或创造价值增量的New technologies are developed to solve problems or to createincremental value. 智能驾驶大模型应用Applications of the foundation model on autonomous driving 智能驾驶核心要素Key Elements of Intelligent Driving 安全：安全≠安全感 ◼智能驾驶长尾效应带来的安全困境Safety Dilemma Caused by the Long Tail Effect ofIntelligent Driving.◼感知大多数还停留在标注阶段，缺少认识能力Most perception remains in the annotation stage, lackingcognitive ability. ◼大量冗余传感器及技术，系统成本居高不下 High system costs due to redundant sensors and technologies.◼大规模的数据采集标注、软硬件设计开发Large-scale data collection, annotation, software, and hardwaredesign and development. High takeover rate in complex road and congested trafficconditions.◼智能驾驶体验未实现全驾驶场景覆盖，体验不连贯Incomplete coverage of driving scenarios in intelligentdriving experience, leading to inconsistent experiences. D a t a-d r i v e n m o d e l i t e r a t i o n a n d e x p e r i e n c e e n h a n c e m e n t 智能驾驶大模型应用Applications of the foundation model on autonomous driving 智能驾驶大模型应用Applications of the foundation model on autonomous driving 智能驾驶大模型应用Applications of the foundation model on autonomous driving 将数字孪生技术应用在自动驾驶研发测试上，在虚拟空间中建立物理世界模型，还原真实世界道路场景、交通流，构建元宇宙智驾仿真技术平台，应用车辆动力学建模和物理级传感器建模关键技术和自动标注功能模块，高效合成标注数据，实现自动驾驶算法数据训练，让数据驱动更安全的自动驾驶。Applying digital twin technology to autonomous driving research and testing, establishing physical world models in virtual space, reconstructing real-world road scenes and traffic flow, building a meta-universe intelligent driving simulation technology platform, and applying key technologies such as vehicle dynamics modeling andphysical-level sensor modeling and automatic labeling function modules, efficiently synthesizing annotated data, achieving data-driven safer autonomous driving. 智能驾驶大模型应用Applications of the foundation model on autonomous driving Performance of Sim2Real style transfer technology ◼实验g相对a提升3.17%；加入迁移数据后，16个类别中的15个类别得到提升；◼Experiment gimproved by 3.17%compared to a; after adding the Sim2Real data, 15 out of 16 categories were improved. 智能驾驶大模型应用Applications of the foundation model on autonomous driving ◼Backbone选用图文多模态模型，大大增强了模型的理解和泛化能力；The backbone adopts the multimodal model of text and images, which greatly enhances the understanding and generalization ability; ◼结合多方数据源，及数据强化策略使模型更好地泛化业务场景;Combining multiple data sources and data enhancement strategies allows the model to fit better in different scenarios;◼最先进的多尺度特征、去噪训练等策略的引入使得Transformer架构性能优越;The introduction of multi-scale features, denoising training, and other strategies makes the Transformer superior in performance;◼可同时处理语义分割、物体检测等2D图像感知任务，可实现标注数据互通。It can processing 2D image perception tasks such as semantic segmentation and object detection simultaneously, and realize annotation datainteroperability. 智能驾驶大模型应用Applications of the foundation model on autonomous driving 预标注大模型效果Performance of Pre-labeling foundation model 国际数据集International data set 以上是在模型在国际数据集Cityscapes、ACDC以及量产回传数据上的推理效果图；The above shows the inference results on international datasets such as Cityscapes, ACDC, and production return data. 智能驾驶大模型应用Applications of the foundation model on autonomous driving 大模型赋能数据合成技术Foundation model empowers data synthesis technology ACDC：极端天气场景数据集ACDC：Adverse Conditions Dataset withCorrespondences CityScapes：语义风格的标杆数据集CityScapes：The benchmark dataset for semantic segmentation 基于AI大模型与虚拟数据合成技术，吉利汽车在国际知名数据集CityScapes（语义风格的标杆数据集）及ACDC数据集（极端天气场景数据集）上，取得了实时榜单全球第一的成绩。 Based on AI foundation model and virtual data synthesis technology,GeelyAutomobile ink. rankedthe first place inthe real-time liston the internationally renowned datasetCityScapesandACDC. 智能座舱大模型应用Foundationmodels applied to intelligent cockpit 音 乐 不 仅 可 以 被 听 见，还 可 以 被 看 见M u s i cc a n n o t o n l y b e h e a r d b u t a l s o s e e n 银河E8大模型音乐律动GEELY Galaxy E8 FoundationModel-Musical Rhythm 自然语言对话+更多…N a t u r a ll a n g u a g ed i a l o g u e+m o r e. . . 大模型研发底座：吉利星睿智算中心 FoundationmodelR&D base：Geely Xingrui lntelligentComputingCenter 基 于 英 伟 达 算 力 工 具，构 建 全 球 车 企 首 个“云-数-智”一 体 化 超 级 云 计 算 平 台 Building the world's first "cloud digital intelligence" integrated super clo