Geometry-Aware Recurrent Neural Networks for Active Visual Recognition

Abstract

We present recurrent geometry-aware neural networks that integrate visual information across multiple views of a scene into 3D latent feature tensors, while maintaining an one-to-one mapping between 3D physical locations in the world scene and latent feature locations. Object detection, object segmentation, and 3D reconstruction is then carried out directly using the constructed 3D feature memory, as opposed to any of the input 2D images. The proposed models are equipped with differentiable egomotion-aware feature warping and (learned) depth-aware unprojection operations to achieve geometrically consistent mapping between the features in the input frame and the constructed latent model of the scene. We empirically show the proposed model generalizes much better than geometryunaware LSTM/GRU networks, especially under the presence of multiple objects and cross-object occlusions. Combined with active view selection policies, our model learns to select informative viewpoints to integrate information from by "undoing" cross-object occlusions, seamlessly combining geometry with learning from experience.

To appear in NIPS2018

Cited by in corpus (5)

• Learning 3D Dynamic Scene Representations for Robot Manipulation
• Active 6D Multi-Object Pose Estimation in Cluttered Scenarios with Deep Reinforcement Learning
• Environment Predictive Coding for Embodied Agents
• MVS2D: Efficient Multi-view Stereo via Attention-Driven 2D Convolutions
• GroundNet: Monocular Ground Plane Normal Estimation with Geometric Consistency