Search Results for "self-extracing"

...To aggregate spatial information, we design spatial cross-attention that each BEV query extracts the spatial features from the regions of interest across camera views. For temporal information, we propose temporal self-attention to recurrently fuse the history BEV information. Our approach achieves the new state-of-the-art 56.9\% in terms of NDS metric on the nuScenes \texttt{test} set, which is 9.0 points higher than previous best arts and on par with the performance of LiDAR-based baseline.

Autonomous driving is gaining increasing attention in the computer vision research community, as vision based scene understanding is key to self-driving cars. In this web page, we make image datasets public for the purpose of furthering research in scene understanding.