Journals / Bilişim Teknolojileri Dergisi / 2019 / Cilt: 12 - Sayı: 4

Labeling System with EEG, EMG, and IMU for Visual Training of Autonomous Vehicles

Otonom Araçların Görsel Eğitimi için EEG, EMG ve IMU ile Etiketleme Sistemi

Pages
299–305
Publication Date
2019-10-29
DOI
—
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

According to the preliminary study findings, the K-NN method was the most successful algorithm among the algorithms tested with 0.922 accuracy.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:Autonomous vehicles are tools that make decisions and take decisions by perceiving their environment.
📊
Key Findings & Data:In the time and frequency field, attributes were extracted by using the non-overlapping sliding window with 0.33 sec length.
💡
Conclusion & Impact:According to the preliminary study findings, the K-NN method was the most successful algorithm among the algorithms tested with 0.922 accuracy.
🔬Design:Modeling & Simulation
📍Scope:Comprehensive Sample
⚙️Variables:EEG · EMG · IMU

Abstract

Autonomous vehicles are tools that make decisions and take decisions by perceiving their environment. Today, autonomous vehicles are also used in traffic in some countries. Various types of cameras, laser radars (LIDAR), sonar distance sensors, etc. are used for environmental detection in autonomous vehicles. After the environment is perceived, the collected data is taught to the vehicle with the help of machine learning methods and the vehicle reaches the target by following the traffic rules. At the point of traffic rules, the biggest task belongs to image-based systems. However, ideal traffic conditions and environmental conditions are not always provided. It is important to identify situations that may present a danger to autonomous vehicles. When the literature is examined, no visual data set or a scientific study with dangerous labeling has been found. In this study, it is aimed to design a data collection and labeling system to overcome this gap in the literature. In the system designed for the purpose, a system which automatically creates a video label from the physiological data of the driver (EEG ve EMG) and the inertia change data during human driving is designed. For this reason, firstly, the sensor signals were collected by experiments. In the time and frequency field, attributes were extracted by using the non-overlapping sliding window with 0.33 sec length. The input variables in the data set were reduced by PCA and classified by DT, RF and K-NN algorithms. According to the preliminary study findings, the K-NN method was the most successful algorithm among the algorithms tested with 0.922 accuracy.

Özet

Otonom araçlar, çevrelerini algılayarak kararlar alan ve bu kararlar ile hareket eden araçlardır. Günümüzde otonom araçlar bazı ülkelerde trafikte de kullanılmaktadır. Otonom araçlarda çevre algılama için çeşitli kameralar, lazer radarlar (LIDAR), sonar sensörler vb. pek çok sensör kullanılmaktadır. Çevre algılandıktan sonra toplanan veri makine öğrenmesi yöntemleri yardımıyla araca öğretilmekte ve araç trafik kurallarına uyarak hedefe ulaşmaktadır. Trafik kuralları noktasında en büyük görev görüntü tabanlı sistemlere düşmektedir. Ancak ideal trafik koşulları ve çevre şartları her zaman sağlanamamaktadır. Bu nedenle otonom araçlar için tehlike oluşturabilecek durumların tespiti önem arz etmektedir. Literatür incelendiğinde tehlikeli durumların etiketli bulunduğu görsel bir veri seti veya ilgili bir bilimsel çalışmaya rastlanmamıştır. Bu çalışmada literatürdeki açığı gidermek için bir veri toplama ve etiketleme sistemi tasarlanması amaçlanmıştır. Amaç doğrultusunda tasarlanan sistemde insan sürüşü esnasında sürücünün fizyolojik verisi (EEG ve EMG) ve eylemsizlik değişim verilerinden otomatik olarak video etiketi oluşturan bir sistem tasarlanmıştır. Bunun için öncelikle deneyler ile sensör sinyalleri toplanmıştır. Toplanan sinyallerden 0.33 sn uzunluğunda üst üste binmeyen kayan pencere kullanılarak zaman ve frekans alanında öznitelikler çıkarılmıştır. Elde edilen veri setindeki giriş değişkenleri Temel Bileşen Analizi (PCA) ile indirgenmiş ve Karar Ağacı (DT), Rastgele Ağaç (RF) ve K En Yakın Komşular (K-NN) algoritmaları ile sınıflandırma işlemine tutulmuştur. Bulgulara göre K-NN yönteminin 0.922 doğrulukla tehlikeli, tehlikesiz durumları ayırt ederek denenen algoritmalar arasında en başarılı algoritma olduğu tespit edilmiştir.

🔖

Cite This Article

SEÇKİN, A. Ç. (2019). Labeling System with EEG, EMG, and IMU for Visual Training of Autonomous Vehicles. Bilişim Teknolojileri Dergisi, Cilt: 12 - Sayı: 4, 299–305. https://ajindex.com/articles/251814

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (18)

  1. B. Siciliano, O. Khatib, Springer Handbook of Robotics, Springer, 2016
  2. D. Nistér, O. Naroditsky, J. Bergen, “Visual Odometry for Ground Vehicle Applications”, Journal of Field Robotics, 23(1), 3–20, 2006
  3. M. O. Aqel, M. H. Marhaban, M. I. Saripan, N. B. Ismail, “Review of visual odometry: types, approaches, challenges, and applications”, SpringerPlus, 5(1), 1897, 2016
  4. H. Lum, J. A. Reagan, “Interactive highway safety design model: accident predictive module”, Public Roads, 58(3), 1995
  5. T. Bliss, Implementing the recommendations of the world report on road traffic injury prevention, 2004
  6. M. Peden et al., World report on road traffic injury prevention, World Health Organization Geneva, 2004
  7. N. J. Goodall, “Can you program ethics into a self-driving car?”, IEEE Spectrum, 53(6), 28–58, 2016
  8. D. Birnbacher, W. Birnbacher, “Fully autonomous driving: Where technology and ethics meet”, IEEE Intelligent Systems, 32(5), 3–4, 2017
  9. N. A. Greenblatt, “Self-driving cars and the law”, IEEE spectrum, 53(2), 46–51,2016
  10. J. Fleetwood, “Public health, ethics, and autonomous vehicles”, American journal of public health, 107(4), 532–537,2017
  11. K. Sadeghi, A. Banerjee, J. Sohankar, S. K. Gupta, “Safedrive: An autonomous driver safety application in aware cities”, Pervasive Computing and Communication Workshops (PerCom Workshops), 2016 IEEE International Conference on, 1–6, 2016
  12. C. Richter, N. Roy, “Safe visual navigation via deep learning and novelty detection”, 2017
  13. D. Gruyer, V. Magnier, K. Hamdi, L. Claussmann, O. Orfila, A. Rakotonirainy, “Perception, information processing and modeling: Critical stages for autonomous driving applications”, Annual Reviews in Control, 44, 323–341,2017
  14. N. Smolyanskiy, A. Kamenev, J. D. Smith, S. T. Birchfield, “Performing autonomous path navigation using deep neural networks”, Oct-2018
  15. L. Chi, Y. Mu, “Deep steering: Learning end-to-end driving model from spatial and temporal visual cues”, arXiv preprint arXiv:1708.03798, 2017
  16. A. Bemporad, A. De Luca, G. Oriolo, “Local incremental planning for a car-like robot navigating among obstacles”, Proceedings of IEEE International Conference on Robotics and Automation, 2, 1205–1211, 1996
  17. J. Liu, P. Jayakumar, J. L. Stein, T. Ersal, “A study on model fidelity for model predictive control-based obstacle avoidance in high-speed autonomous ground vehicles”, Vehicle System Dynamics, 54(11), 1629–1650,2016
  18. N. Nourani-