# Recurrent Scene Parsing with Perspective Understanding in the Loop

## TL;DR

This paper introduces a depth-aware gating module integrated into a recurrent CNN for semantic segmentation, adaptively adjusting pooling based on object depth to improve accuracy and efficiency in perspective images.

## Contribution

It presents a novel depth-aware gating mechanism that dynamically adjusts receptive fields in CNNs, enhancing segmentation performance with a more compact model.

## Key findings

- Achieves competitive semantic segmentation results on large-scale RGB-D datasets.
- State-of-the-art monocular depth estimation using gated pooling.
- Demonstrates effectiveness of depth as side-information during training.

## Abstract

Objects may appear at arbitrary scales in perspective images of a scene, posing a challenge for recognition systems that process images at a fixed resolution. We propose a depth-aware gating module that adaptively selects the pooling field size in a convolutional network architecture according to the object scale (inversely proportional to the depth) so that small details are preserved for distant objects while larger receptive fields are used for those nearby. The depth gating signal is provided by stereo disparity or estimated directly from monocular input. We integrate this depth-aware gating into a recurrent convolutional neural network to perform semantic segmentation. Our recurrent module iteratively refines the segmentation results, leveraging the depth and semantic predictions from the previous iterations. Through extensive experiments on four popular large-scale RGB-D datasets, we demonstrate this approach achieves competitive semantic segmentation performance with a model which is substantially more compact. We carry out extensive analysis of this architecture including variants that operate on monocular RGB but use depth as side-information during training, unsupervised gating as a generic attentional mechanism, and multi-resolution gating. We find that gated pooling for joint semantic segmentation and depth yields state-of-the-art results for quantitative monocular depth estimation.