To evaluate our model, we develop AutoCastSim, a network-augmented driving simulation framework with example accident-prone scenarios.
We propose a novel way to leverage self-supervised contrastive learning to facilitate target feature learning, along with an online pseudo labeling scheme with refinement that significantly denoises pseudo labels.
Experiments on the KITTI and DDAD datasets show that our DepthFormer architecture establishes a new state of the art in self-supervised monocular depth estimation, and is even competitive with highly specialized supervised single-frame architectures.
In this paper, we present a system to train driving policies from experiences collected not just from the ego-vehicle, but all vehicles that it observes.
Ranked #1 on Autonomous Driving on CARLA Leaderboard
Active learning for object detection is conventionally achieved by applying techniques developed for classification in a way that aggregates individual detections into image-level selection criteria.
In deep learning, a typical strategy for transfer learning is to freeze the early layers of a pre-trained model and fine-tune the rest of its layers on the target domain.
This assumption greatly simplifies the learning problem, factorizing the dynamics into a nonreactive world model and a low-dimensional and compact forward model of the ego-vehicle.
Ranked #5 on Autonomous Driving on CARLA Leaderboard
Active learning aims to develop label-efficient algorithms by querying the most representative samples to be labeled by a human annotator.
The agent uses its current segmentation model to infer pixels that constitute objects and refines the segmentation model by interacting with these pixels.
In our framework, the role of the expert is only to communicate the goals (i. e., what to imitate) during inference.
Manipulation of deformable objects, such as ropes and cloth, is an important but challenging problem in robotics.