We're thrilled to announce DeepRoute.ai's participation in The European Conference on Computer Vision (ECCV) 2024, which is a biennial premier research conference in Computer Vision and Machine Learning. At this year's ECCV, our team will present two papers that advance end-to-end autonomous driving technology. We will host a poster presentation and livestream paper reading of the first topic.

• PPAD: Iterative Interactions of Prediction and Planning for End-to-end Autonomous Driving.

  Our team presents a new interaction mechanism of prediction and planning for end-to-end autonomous driving, called PPAD (Iterative Interaction of Prediction and Planning Autonomous Driving), which considers the timestep-wise interaction to better integrate prediction and planning. An ego vehicle performs motion planning at each timestep based on the trajectory prediction of surrounding agents (e.g., vehicles and pedestrians) and its local road conditions. PPAD models the interactions among ego, agents, and the dynamic environment in an autoregressive manner by interleaving the Prediction and Planning processes at every timestep, instead of a single sequential process of prediction followed by planning. The experiments on the nuScenes benchmark show that our approach outperforms state-of-the-art methods. (https://arxiv.org/html/2311.08100v3)

  
  

  
  

• Learning High-resolution Vector Representation from Multi-Camera Images for 3D Object Detection.

  Traditional BEV grid representation induces quadratic computational cost as the spatial resolution grows. We present a new camera-based 3D object detector with high-resolution vector representation: VectorFormer. The presented high-resolution vector representation is combined with the lower-resolution BEV representation to efficiently exploit 3D geometry from multi-camera images at a high resolution through our two novel modules: vector scattering and gathering. Thus, the learned vector representation with richer scene contexts can serve as the decoding query for final predictions. 

Meet DeepRoute.ai team at ECCV

If you're attending ECCV in person, you will find our team at poster presentation. If you're interested in the paper reading of end-to-end autonomous driving, you can join with following details.

Voov Link: https://meeting.tencent.com/dm/3wqPVHJI67CH

*Please note that paper reading will be conducted in Chinese.

We look forward to seeing you in Milan at ECCV 2024!