Abstract
Object instance segmentation is a challenging task in computer vision research, and instance annotation is its key sub-task and a basic component of segmentation models. In this paper, we propose the Deep Active Curve Network (DACN) which combines powerful ResNet models with GCN-based active curves. This new method regards instance segmentation as a points/vertices, edges and object masks prediction task instead of only a pixel-labeling problem. The multi-scale encoder firstly predicts a coarse result through a combination of the edge and segmentation features in a multi-task learning framework, which is effective dealing with objects with rough boundaries. In order to generate an accurate object annotation, an iterative Graph Convolutional Network (GCN) is used to correct the encoder’s feature map and move all vertices of the predicted coarse result to the edge of the corresponding ground truth. A novel weighted loss function further estimates the location of points, edges, and segmented areas, and optimizes the annotation results. Finally, we use the improved 5-interpolation Catmull-Rom spline (CRS) algorithm which exploits key points to control the active curves around objects. In the experimental analysis, we demonstrate the effectiveness of our proposed method on three datasets in automatic mode, including Cityscapes, ADE20K and Rooftop. We further show the generalization ability of our approach on two novel cross-domain datasets.
Highlights
Object instance segmentation is an important task that predicts the detail structures of a given object with a specific category in an image
The encoder feature and an initialization with point sets that make up a circular curve are decoded and corrected by an iterative Graph Convolutional Network (GCN); these predicted key vertices are controlled by the active curve algorithm to generate the final refine prediction
EXPERIMENTAL ANALYSIS we evaluate our approach for the task of object instance annotation and segmentation on the Cityscapes dataset, and provide additional results on cross-domain datasets such as ADE20K [43] and Rooftop [6] to verify the generalizability of our method
Summary
Object instance segmentation is an important task that predicts the detail structures of a given object with a specific category in an image. It is crucial in the field of intelligent traffic monitoring and autonomous driving to locate and segment moving objects such as cars, pedestrians, and other transportations [1], [2]. In addition to the transportation research, automatically extracting building footprints from aerial imagery is widely used in remote sensing [3]–[6]. These applications motivate the development of automatic instance annotation and segmentation to form a clear understanding of an object or region.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have