Safety Prediction for Autonomous Vehicles Using Artificial Intelligence with Modified Deep CNN–BiLSTM
Dr.Sophiya Sanjay Bartalwar
In the recent world, the usages of autonomous cars are getting higher because of the emerging technology. People consider that these autonomous cars make their travel safe in any type of road conditions. These autonomous cars give freedom to drive the person those who are not able to drive. It can able to control the CO2 gas emission, avoid traffic and accidents and there are no attention issues like human in autonomous cars. However, the autonomous cars are not perfect because sometimes the autonomous cars face some issues while analysing the different human hand gesture, climatic conditions and road sign. Due to drivers’ health condition and malfunctions in the autonomous cars also met with an accident. Some manual techniques are used to analyse the drivers’ health to ensure the safety such as survey, dashboard camera and so on. However, these techniques need some time and manual work. To overcome this problem the proposed model use improved search ability based GA (Genetic Algorithm) in feature selection to attain the best features from the dataset and to predict the drivers behaviour and car mechanism the modified deep CNN (Convolutional Neural Network) –BiLSTM (Bidirectional Long Short Term Memory) algorithm with attention mechanism is used. The CNN method is not able to handle the time series data but it is faster and the BiLSTM is able to handle the time series data. Therefore, the proposed model integrated the CNN and BiLSTM to attain best performance. The model included DT (Decision Tree), KNN, NB (Naive Bayes), RF (Random Forest) along with modified deep CNN-BiLSTM with AM (Attention Mechanism). The proposed method attains high performance accuracy while compared with other models.
"Safety Prediction for Autonomous Vehicles Using Artificial Intelligence with Modified Deep CNN–BiLSTM ", IJSDR - International Journal of Scientific Development and Research (www.IJSDR.org), ISSN:2455-2631, Vol.11, Issue 1, page no.b245-b312, January-2026, Available :https://ijsdr.org/papers/IJSDRTH01016.pdf
Volume 11
Issue 1,
January-2026
Pages : b245-b312
Paper Reg. ID: IJSDR_307053
Published Paper Id: IJSDRTH01016
Downloads: 000393
Research Area: Science and Technology
Country: Bhilai, chhattisgarh, India
ISSN: 2455-2631 | IMPACT FACTOR: 9.15 Calculated By Google Scholar | ESTD YEAR: 2016
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 9.15 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator
Publisher: IJSDR(IJ Publication) Janvi Wave
