Search papers, labs, and topics across Lattice.
3
0
2
Train high-speed, event-camera-guided quadrotors without expensive event rendering by decoupling representation learning from policy learning.
Suppressing motion-related noise in event camera data can boost Vision Transformer inference by 83% and improve visual odometry accuracy by 13%.
By fusing asynchronous event streams with motion-blurred images, this method achieves a 50% performance boost in NeRF reconstruction for fast-flying drones, even when pose priors are unreliable.
