Multi-stream CNN based Video Semantic Segmentation for Automated Driving

被引:3
|
作者
Sistu, Ganesh [1 ]
Chennupati, Sumanth [2 ]
Yogamani, Senthil [1 ]
机构
[1] Valeo Vis Syst, Dublin, Ireland
[2] Valeo Troy, Troy, NY USA
关键词
Semantic Segmentation; Visual Perception; Automated Driving;
D O I
10.5220/0007248401730180
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Majority of semantic segmentation algorithms operate on a single frame even in the case of videos. In this work, the goal is to exploit temporal information within the algorithm model for leveraging motion cues and temporal consistency. We propose two simple high-level architectures based on Recurrent FCN (RFCN) and Multi-Stream FCN (MSFCN) networks. In case of RFCN, a recurrent network namely LSTM is inserted between the encoder and decoder. MSFCN combines the encoders of different frames into a fused encoder via 1x1 channel-wise convolution. We use a ResNet50 network as the baseline encoder and construct three networks namely MSFCN of order 2 & 3 and RFCN of order 2. MSFCN-3 produces the best results with an accuracy improvement of 9% and 15% for Highway and New York-like city scenarios in the SYNTHIA-CVPR'16 dataset using mean IoU metric. MSFCN-3 also produced 11% and 6% for SegTrack V2 and DAVIS datasets over the baseline FCN network. We also designed an efficient version of MSFCN-2 and RFCN-2 using weight sharing among the two encoders. The efficient MSFCN-2 provided an improvement of 11% and 5% for KITTI and SYNTHIA with negligible increase in computational complexity compared to the baseline version.
引用
收藏
页码:173 / 180
页数:8
相关论文
共 50 条
  • [31] Medical image segmentation based on active fusion-transduction of multi-stream features?
    Shu, Yucheng
    Zhang, Jing
    Xiao, Bin
    Li, Weisheng
    KNOWLEDGE-BASED SYSTEMS, 2021, 220
  • [32] Arabic Handwriting Recognition Based on Synchronous Multi-stream HMM Without Explicit Segmentation
    Jayech, Khaoula
    Mahjoub, Mohamed Ali
    Ben Amara, Najoua Essoukri
    HYBRID ARTIFICIAL INTELLIGENT SYSTEMS (HAIS 2015), 2015, 9121 : 136 - 145
  • [33] Multi-stream Attention-based BLSTM with Feature Segmentation for Speech Emotion Recognition
    Chiba, Yuya
    Nose, Takashi
    Ito, Akinori
    INTERSPEECH 2020, 2020, : 3301 - 3305
  • [34] Automated speech recognition by multi-stream dynamic time warping
    Mohamadi, T
    Gharbi, AH
    Mezaache, S
    Harrag, A
    CANADIAN CONFERENCE ON ELECTRICAL AND COMPUTER ENGINEERING 2001, VOLS I AND II, CONFERENCE PROCEEDINGS, 2001, : 527 - 531
  • [35] Stochastic Fusion for Multi-stream Neural Network in Video Classification
    Huang, Yu-Min
    Tseng, Huan-Hsin
    Chien, Jen-Tzung
    2019 ASIA-PACIFIC SIGNAL AND INFORMATION PROCESSING ASSOCIATION ANNUAL SUMMIT AND CONFERENCE (APSIPA ASC), 2019, : 69 - 74
  • [36] BiSwift: Bandwidth Orchestrator for Multi-Stream Video Analytics on Edge
    Su, Lin
    Wang, Weijun
    Yuan, Tingting
    Li, Liang
    Dai, Aipeng
    Liu, Yunxin
    Fu, Xiaoming
    IEEE INFOCOM 2024-IEEE CONFERENCE ON COMPUTER COMMUNICATIONS, 2024, : 1181 - 1190
  • [37] Multi-stream video transport over DiffServ wireless LANs
    Man, H
    Li, Y
    2003 INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, VOL II, PROCEEDINGS, 2003, : 541 - 544
  • [38] On Adversarial Robustness of Semantic Segmentation Models for Automated Driving
    Yin, Huilin
    Wang, Ruining
    Liu, Boyu
    Yan, Jun
    2022 IEEE INTELLIGENT VEHICLES SYMPOSIUM (IV), 2022, : 867 - 873
  • [39] Automated Evaluation of Semantic Segmentation Robustness for Autonomous Driving
    Zhou, Wei
    Berrio, Julie Stephany
    Worrall, Stewart
    Nebot, Eduardo
    IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2020, 21 (05) : 1951 - 1963
  • [40] Leveraging Semantic Scene Characteristics and Multi-Stream Convolutional Architectures in a Contextual Approach for Video-Based Visual Emotion Recognition in the Wild
    Pikoulis, Ioannis
    Filntisis, Panagiotis P.
    Maragos, Petros
    2021 16TH IEEE INTERNATIONAL CONFERENCE ON AUTOMATIC FACE AND GESTURE RECOGNITION (FG 2021), 2021,