Loop Closure Detection Algorithm Based on Mixed Global Pooling

doi:10.3969/j.issn.1006-2475.2020.04.019

Abstract

Abstract: The deep learning based loop closure detection algorithm has been verified to be superior to traditional methods. However, the computation burden of deep learning is heavy, so it is often difficult to deploy large convolutional neural networks on mobile robots, while small convolutional neural networks perform poorly in large scenes. Therefore, this paper proposes a scheme to deploy large convolutional neural networks on mobile robots. Firstly, the feature graph is transformed into the feature vector by using the mixed global pooling layer. Experiments show that the performance of this method is equivalent to that of other more complex methods and the calculation is simpler. Then, a block-based floating-point convolutional neural network acceleration engine is proposed, which significantly reduces the computational energy consumption and causes almost no performance loss without retraining.

Key words: Key words: visual SLAM, loop closure, deep learning, CNN accelerator, mobile robot

CLC Number:

TP399

SONG Zhou-rui. Loop Closure Detection Algorithm Based on Mixed Global Pooling[J]. Computer and Modernization, doi: 10.3969/j.issn.1006-2475.2020.04.019.

References

［1］ FILIP R, GIORGOS T, ONDREJ C. Fine-tuning CNN image retrieval with no human annotation［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018,41(7):1655-1668.
［2］ KAVITHA M L, DEEPAMBIKA V A. Review on loop closure detection of visual SLAM［J］. International Journal of Current Engineering and Scientific Research, 2018,6(6):81-86.
［3］ BABENKO A V, VICTOR L. Aggregating local deep features for image retrieval［C］// 2016 IEEE International Conference on Computer Vision. 2016:1269-1277.
［4］ TOLIAS G, SICRE R, JGOU H. Particular object retrieval with integral max-pooling of CNN activations［J］. Computer Science, 2015,1(1):1-18.
［5］ BABENKO A, SLESAREV A, CHIGORIN A, et al. Neural codes for image retrieval［C］// 2014 European Conference on Computer Vision. 2014:584-599.
［6］ BAI D D, WANG C Q, ZHANG B, et al. CNN feature boosted SeqSLAM for real-time loop closure detection［J］. Chinese Journal of Electronics, 2018,27(3):488-499.
［7］ GAO X, ZHANG T. Unsupervised learning to detect loops using deep neural networks for visual SLAM system［J］. Autonomous Robots, 2017,41(1):1-18.
［8］ MERRILL N, HUANG G. Lightweight unsupervised deep loop closure［J］. Robotics, 2018:arXiv:1805.07703.
［9］ LIU H, ZHAO C Y, HUANG W P, et al. An end-to-end Siamese convolutional neural network for loop closure detection in visual slam system［C］// International Conference on Acoustics, Speech, and Signal Processing. 2018:3121-3125.
［10］CHEN Y H, KRISHNA T, EMER J S, et al. Eyeriss: An energy-efficient reconfigurable accelerator for deep convolutional neural networks［J］. IEEE Journal of Solid-State Circuits, 2017,52(1):127-138.
［11］ZHANG C, LI P, SUN G Y, et al. Optimizing FPGA-based accelerator design for deep convolutional neural networks［C］// Proceedings of 2015 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays. 2015:161-170.
［12］HAN S, MAO H Z, DALLY W J. Deep compression: Compressing deep neural networks with pruning, trained quantization and huffman coding［J］. Computer Vision and Pattern Recognition, 2015:arXiv:1510.00149.
［13］HAN S, POOL J, TRAN J, et al. Learning both weights and connections for efficient neural network［C］// Advances in Neural Information Processing Systems. 2015:1135-1143.
［14］PAGE A, MOHSENIN T. FPGA-based reduction techniques for efficient deep neural network development［C］// 2016 IEEE 24th Annual International Symposium on Field-Programmable Custom Computing Machines. 2016:200.
［15］GUPTA S, AGRAWAL A, GOPALAKRISHNAN K, et al. Deep learning with limited numerical precision［C］// 2015 International Conference on Machine Learning. 2015:1737-1746.
［16］HILL P, ZAMIRAI B, LU S, et al. Rethinking numerical representations for deep neural networks［J］. Machine Learning, 2018:arXiv:1808.02513.
［17］MELLEMPUDI N, KUNDU A, DAS D, et al. Mixed low-precision deep learning inference using dynamic fixed point［J］. Machine Learning, 2017:arXiv:1701.08978.
［18］MEI C S, LIU Z Y, NIU Y, et al. A 200MHz 202.4 GFLOPS@ 10.8 W VGG16 accelerator in Xilinx VX690T［C］// 2017 IEEE Global Conference on Signal and Information Processing. 2017:784-788.
［19］KALLIOJARVI K, ASTOLA J. Roundoff errors in block-floating-point systems［J］. IEEE Transactions on Signal Processing, 1996,44(4):783-790.
［20］SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition［J］. Computer Vision and Pattern Recognition, 2014:arXiv:1409.1556.
［21］LEE C Y, GALLAGHER P W, TU Z. Generalizing pooling functions in convolutional neural networks: Mixed, gated, and tree［C］// Artificial Intelligence and Statistics. 2016:464-472.
［22］JIA Y Q, SHELHAMER E, DONAHUE J, et al. Caffe: Convolutional architecture for fast feature embedding［C］// Proceedings of the 22nd ACM International Conference on Multimedia. 2014:675-678.
［23］DENG J, DONG W, SOCHER R, et al.ImageNet: A large-scale hierarchical image database［C］// 2009 IEEE Conference on Computer Vision and Pattern Recognition. 2009:248-255.
［24］KALANTIDIS Y, MELLINA C, OSINDERO S. Cross-dimensional weighting for aggregated deep convolutional features［C］// European Conference on Computer Vision. 2016:685-701.

[1]	HU Chong-jia, LIU Jin-zhou, FANG Li. Unsupervised Domain Adaptation for Outdoor Point Cloud Semantic Segmentation [J]. Computer and Modernization, 2024, 0(01): 74-79.
[2]	LIN Wei. Incremental News Recommendation Method Based on Self-supervised Learning and Data Replay [J]. Computer and Modernization, 2023, 0(12): 1-6.
[3]	LIANG Tian-kai, HUANG Kang-hua, LIU Kai-hang, LAN Lan, ZENG Bi. Deep Federated Image Classification Method Based on Bilateral Homomorphic Encryption [J]. Computer and Modernization, 2023, 0(12): 36-40.
[4]	QIU Kai-xing, FENG Guang. A Multi-label Image Classification Model Based on Dual Feature Attention [J]. Computer and Modernization, 2023, 0(12): 41-47.
[5]	ZHANG Bo-quan, MAI Hai-peng, CHEN Jia-min, Pang Jin-ju. White Matter Hyperintensities Segmentation Based on High Gray Value#br# Attention Mechanism [J]. Computer and Modernization, 2023, 0(12): 67-75.
[6]	LI Yan-man, WANG Bi-heng, ZHAO Ling-yan. Safety Helmet Detection Based on Lightweight YOLOv5 [J]. Computer and Modernization, 2023, 0(10): 59-64.
[7]	LI Shi-da, XIANG Jian-wen. A Weakened Joint Reinforcement Method to Improve Robustness of Image Recognition Models [J]. Computer and Modernization, 2023, 0(10): 70-76.
[8]	SHEN Jia-wei, LU Yi-ming, CHEN Xiao-yi, QIAN Mei-ling, LU Wei-zhong, . Review of Research on Human Behavior Detection Methods Based on Deep Learning [J]. Computer and Modernization, 2023, 0(09): 1-9.
[9]	LIU Chan-yi, HUANG Dan, XUE Lin-yan, WANG Tao, ZHU Tao, . COVID-19 X-ray Classification Based on Improved Efficientnet Network [J]. Computer and Modernization, 2023, 0(09): 94-99.
[10]	MA Guo-xiang, YANG Ling-fei, YAN Chuan-bo, ZHANG Zhi-hao, SUN Bing, WANG Xiao-rong. Ultrasonic Image Diagnosis of Hepatic Echinococcosis Based on Deep DenseNet Network [J]. Computer and Modernization, 2023, 0(09): 100-104.
[11]	NONG Hao-cheng, REN De-jun, REN Qiu-lin, LIU Peng-li, HUANG De-cheng. Surface Anomaly Detection Algorithm of Flexible Plastic Packaging Based on Improved ConvNeXt [J]. Computer and Modernization, 2023, 0(08): 12-17.
[12]	OUYANG Fei, WU Xu, XIANG Dong-sheng. Garbage Classification and Detection Method Based on Improved YOLOX [J]. Computer and Modernization, 2023, 0(08): 68-73.
[13]	HU Rui-jie, CHE Dou. Review of Infrared Small Target Detection [J]. Computer and Modernization, 2023, 0(08): 79-86.
[14]	JIANG Lei, TANG Jian, YANG Chao-yue, LYU Ting-ting. Bearing Fault Diagnosis Based on CWGAN-GP and CNN [J]. Computer and Modernization, 2023, 0(07): 1-6.
[15]	WANG Jia-chen, ZHANG Hong-xin, LIU Qing-hua, . CT Image Generation of Pneumonia Based on Generative Adversarial Network [J]. Computer and Modernization, 2023, 0(07): 20-24.