Small Obstacle Avoidance Based on RGB-D Semantic Segmentation

  title={Small Obstacle Avoidance Based on RGB-D Semantic Segmentation},
  author={Minjie Hua and Yibing Nan and Shiguo Lian},
  journal={2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)},
  • M. Hua, Yibing Nan, Shiguo Lian
  • Published 2019
  • Computer Science
  • 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW)
This paper presents a novel obstacle avoidance system for road robots equipped with RGB-D sensor that captures scenes of its way forward. The purpose of the system is to have road robots move around autonomously and constantly without any collision even with small obstacles, which are often missed by existing solutions. For each input RGB-D image, the system uses a new two-stage semantic segmentation network followed by the morphological processing to generate the accurate semantic map… Expand
Can we cover navigational perception needs of the visually impaired by panoptic segmentation?
It is demonstrated that panoptic segmentation is able to equip the visually impaired with a holistic real-world scene perception through a wearable assistive system. Expand
Vision Based Wall Following Framework: A Case Study With HSR Robot for Cleaning Application
A new vision-based wall following framework is proposed that acts as an add-on for any professional robotic platform to perform wall cleaning and could successfully classify and segment the wall/floor surface and also detect the obstacle on wall and floor with high detection accuracy. Expand
Helping the Blind to Get through COVID-19: Social Distancing Assistant Using Real-Time Semantic Segmentation on RGB-D Video
This paper presents a system that helps blind people to maintain physical distance to other persons using a combination of RGB and depth cameras and warns the user only if persons are nearby but does not react to non-person objects such as walls, trees or doors. Expand
FSNet: A Failure Detection Framework for Semantic Segmentation
A failure detection framework is proposed to identify pixel-level misclassification in semantic segmentation by exploiting internal features of the segmentation model and training it simultaneously with a failure detection network. Expand


Color Road Segmentation And Video Obstacle Detection
The primary vision task in road-following for a mobile robot is to provide a description of the road environment, including possible obstacles on the road. Techniques are presented for roadExpand
Robust obstacle segmentation based on topological persistence in outdoor traffic scenes
A new methodology for robust segmentation of obstacles from stereo disparity maps in an on-road environment is presented and performs a topological persistence analysis on the constructed occupancy map to identify the regions that are most persistent. Expand
Appearance-Based Obstacle Detection with Monocular Color Vision
This paper presents a new vision-based obstacle detection method for mobile robots that uses a single passive color camera, performs in real-time, and provides a binary obstacle image at high resolution. Expand
Fast Image-Based Obstacle Detection From Unmanned Surface Vehicles
A new graphical model is proposed that affords a fast and continuous obstacle image-map estimation from a single video stream captured on board a USV, and outperforms the related approaches, while requiring a fraction of computational effort. Expand
Joint perception and planning for efficient obstacle avoidance using stereo vision
This paper introduces an approach to Joint Perception and Planning using stereo vision, which performs disparity checks on demand, only as necessary while searching on a planning graph, and demonstrates that the JPP requires less than 10% of the disparity computations required by SPP. Expand
Visually-guided obstacle avoidance in unstructured environments
  • Liana M. Lorigo, R. Brooks, W. E. L. Grimsou
  • Computer Science
  • Proceedings of the 1997 IEEE/RSJ International Conference on Intelligent Robot and Systems. Innovative Robotics for Real-World Applications. IROS '97
  • 1997
This paper presents an autonomous vision-based obstacle avoidance system. The system consists of three independent vision modules for obstacle detection, each of which is computationally simple andExpand
MergeNet: A Deep Net Architecture for Small Obstacle Discovery
A novel network architecture called MergeNet for discovering small obstacles for on-road scenes in the context of autonomous driving is presented, involving weight-sharing, separate learning of low and high level features from the RGBD input and a refining stage which learns to fuse the obtained complementary features. Expand
Unifying Terrain Awareness for the Visually Impaired through Real-Time Semantic Segmentation
A comprehensive set of experiments prove the qualified accuracy over state-of-the-art methods while maintaining real-time speed and a closed-loop field test involving real visually-impaired users, demonstrating the effectivity and versatility of the assistive framework. Expand
Detecting unexpected obstacles for self-driving cars: Fusing deep learning and geometric modeling
A principled Bayesian framework is presented to fuse the semantic and stereo-based detection results, and the mid-level Stixel representation is used to describe obstacles in a flexible, compact and robust manner. Expand
RedNet: Residual Encoder-Decoder Network for indoor RGB-D Semantic Segmentation
This paper proposes an RGB-D residual encoder-decoder architecture, named RedNet, and proposes a `pyramid supervision' training scheme, which applies supervised learning over different layers in the decoder, to cope with the problem of gradients vanishing. Expand