Loading [MathJax]/extensions/TeX/euler_ieee.js
Interactive Object Segmentation With Inside-Outside Guidance | IEEE Conference Publication | IEEE Xplore

Interactive Object Segmentation With Inside-Outside Guidance

Publisher: IEEE

Abstract:

This paper explores how to harvest precise object segmentation masks while minimizing the human interaction cost. To achieve this, we propose an Inside-Outside Guidance (...View more

Abstract:

This paper explores how to harvest precise object segmentation masks while minimizing the human interaction cost. To achieve this, we propose an Inside-Outside Guidance (IOG) approach in this work. Concretely, we leverage an inside point that is clicked near the object center and two outside points at the symmetrical corner locations (top-left and bottom-right or top-right and bottom-left) of a tight bounding box that encloses the target object. This results in a total of one foreground click and four background clicks for segmentation. The advantages of our IOG is four-fold: 1) the two outside points can help to remove distractions from other objects or background; 2) the inside point can help to eliminate the unrelated regions inside the bounding box; 3) the inside and outside points are easily identified, reducing the confusion raised by the state-of-the-art DEXTR in labeling some extreme samples; 4) our approach naturally supports additional clicks annotations for further correction. Despite its simplicity, our IOG not only achieves state-of-the-art performance on several popular benchmarks, but also demonstrates strong generalization capability across different domains such as street scenes, aerial imagery and medical images, without fine-tuning. In addition, we also propose a simple two-stage solution that enables our IOG to produce high quality instance segmentation masks from existing datasets with off-the-shelf bounding boxes such as ImageNet and Open Images, demonstrating the superiority of our IOG as an annotation tool.
Date of Conference: 13-19 June 2020
Date Added to IEEE Xplore: 05 August 2020
ISBN Information:

ISSN Information:

Publisher: IEEE
Conference Location: Seattle, WA, USA

1. Introduction

Over the past few years, we have witnessed a revolutionary advancement in semantic [44], [40], [68], [69], [10]–[12], [30], [15], [31] and instance segmentation [25], [36], [8], [60], [13], [65], [34], [4], [9], [43], [29] for different domains, such as general scenes [20], [41], [70], autonomous driving [17], [48], [21], aerial imagery [57], [16], medical diagnosis [22], [56], etc. Successful segmentation models are usually built on the shoulders of large volumes of high-quality training data. However, the process to create the pixel-level training data necessary to build these models is often expensive, laborious and time-consuming. Thus, interactive segmentation, which allows the human annotators to quickly extract the object-of-interest by providing some user inputs such as bounding boxes [66], [52], [64] or clicks [67], [38], [45], [37], appears to be an attractive and efficient way to reduce the annotation effort.

(a) User inputs of DEXTR [46]. (b) User inputs of the proposed IOG method. (c) An overview of our IOG framework.

References

References is not available for this document.