Deep Multicameral Decoding for Localizing Unoccluded Object Instances from a Single RGB Image

TL;DR

This paper introduces a novel deep multicameral decoding architecture for localizing unoccluded objects in dense, homogeneous layouts from a single RGB image, along with a synthetic dataset to improve occlusion-aware segmentation.

Contribution

It proposes a multicameral encoder-decoder design with subtask-specific units and introduces Mikado, a synthetic dataset for dense object layouts, enhancing occlusion-aware segmentation performance.

Findings

Multiscale units outperform traditional design patterns.

Mikado dataset enables transfer learning to real images.

Proposed method improves localization in dense occlusion scenarios.

Abstract

Occlusion-aware instance-sensitive segmentation is a complex task generally split into region-based segmentations, by approximating instances as their bounding box. We address the showcase scenario of dense homogeneous layouts in which this approximation does not hold. In this scenario, outlining unoccluded instances by decoding a deep encoder becomes difficult, due to the translation invariance of convolutional layers and the lack of complexity in the decoder. We therefore propose a multicameral design composed of subtask-specific lightweight decoder and encoder-decoder units, coupled in cascade to encourage subtask-specific feature reuse and enforce a learning path within the decoding process. Furthermore, the state-of-the-art datasets for occlusion-aware instance segmentation contain real images with few instances and occlusions mostly due to objects occluding the background, unlike dense object layouts. We thus also introduce a synthetic dataset of dense homogeneous object layouts, namely Mikado, which extensibly contains more instances and inter-instance occlusions per image than these public datasets. Our extensive experiments on Mikado and public datasets show that ordinal multiscale units within the decoding process prove more effective than state-of-the-art design patterns for capturing position-sensitive representations. We also show that Mikado is plausible with respect to real-world problems, in the sense that it enables the learning of performance-enhancing representations transferable to real images, while drastically reducing the need of hand-made annotations for finetuning. The proposed dataset will be made publicly available.