We propose SURF-LSTM, a low complexity deep architecture to learn image absolute pose (position and orientation) in indoor environments using SURF descriptors and recurrent neural networks. Given the strongest SURF features descriptors of an input image, we use 2 layers of bidirectional long short term memory (LSTM) to model the sequential relation between them to learn the 6 degrees of freedom absolute pose in an arbitrary reference frame. In addition to achieving competitive performance compared to existing image localization methods, our system can be trained in less than 10 minutes instead of hours by the state of the art. It requires as small as 0.0128 MB to save the image frame rather than 0.08 MB compared to other methods that use the cropped images and the weights file needs 1.5 MB of storage compared to 100 MB of other methods which leads to significant time and space efficiency.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    SURF-LSTM: A Descriptor Enhanced Recurrent Neural Network For Indoor Localization


    Contributors:


    Publication date :

    2020-11-01


    Size :

    1140318 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    LSTM recurrent neural network assisted aircraft stall prediction for enhanced situational awareness

    Saniat, Tahsin Sejat / Goni, Tahiat / Galib, Shaikat M. | ArXiv | 2020

    Free access

    Short-term traffic flow prediction with LSTM recurrent neural network

    Kang, Danqing / Lv, Yisheng / Chen, Yuan-yuan | IEEE | 2017


    Exploring LSTM based recurrent neural network for failure time series prediction

    Wang, Xin / Wu, Ji / Liu, Chao et al. | British Library Online Contents | 2018


    Vehicle Trajectory Prediction based on LSTM Recurrent Neural Networks

    Ip, Andre / Irio, Luis / Oliveira, Rodolfo | IEEE | 2021