No abstract

Recent techniques in self-supervised monocular depth estimation are approaching the performance of supervised methods, but operate in low resolution only. We show that high resolution is key towards high-fidelity self-supervised monocular depth prediction. Inspired by recent deep learning methods for Single-Image Super-Resolution, we propose a subpixel convolutional layer extension for depth super-resolution that accurately synthesizes high-resolution disparities from their corresponding low-resolution convolutional features. In addition, we introduce a differentiable flip-augmentation layer that accurately fuses predictions from the image and its horizontally flipped version, reducing the effect of left and right shadow regions generated in the disparity map due to occlusions. Both contributions provide significant performance gains over the state-of-the-art in self-supervised depth and pose estimation on the public KITTI benchmark. A video of our approach can be found at https://youtu.be/jKNgBeBMx0I.

Abstract-In this work, we develop a monocular SLAMaware object recognition system that is able to achieve considerably stronger recognition performance, as compared to classical object recognition systems that function on a frame-by-frame basis. By incorporating several key ideas including multi-view object proposals and efficient feature encoding methods, our proposed system is able to detect and robustly recognize objects in its environment using a single RGB camera in near-constant time. Through experiments, we illustrate the utility of using such a system to effectively detect and recognize objects, incorporating multiple object viewpoint detections into a unified prediction hypothesis. The performance of the proposed recognition system is evaluated on the UW RGB-D Dataset, showing strong recognition performance and scalable run-time performance compared to current state-of-the-art recognition systems.

Robots operating autonomously in household environments must be capable of interacting with articulated objects on a daily basis. They should be able to infer each object's underlying kinematic linkages purely by observing its motion during manipulation. This work proposes a framework that enables robots to learn the articulation in objects from user-provided demonstrations, using RGB-D sensors. We introduce algorithms that combine concepts in sparse feature tracking, motion segmentation, object pose estimation, and articulation learning, to develop our proposed framework. Additionally, our methods can predict the motion of previously seen articulated objects in future encounters. We present experiments that demonstrate the ability of our method, given RGB-D data, to identify, analyze and predict the articulation of a number of everyday objects within a human-occupied environment.

No abstract