An-Improved-LSTM-Model-for- Behavior-Recognition-of-Intelligent-Vehicles A total of four experiments have done,including: 1.Vehivle behavior recognition experiment based on vehicle behavior data set from Udacity 2.Handwritten digit recognition experiment based on handwritten number 3.Emotion classification based on movie reviewm 4.Next trajectory point regression fitting experiment based on NGSIM data set.

The vehicle behavior recogntion experiments use the traditional machine learning method,the MLP method and the traditional LSTM classification model and the improved LSTM classification method to do comparative experiments. The other three verification experiments mainly use the traditional LSTM model and the imprioved LSTM classification model to verify the superiority and robustness of improved LSTM classification model.

If you want to know more, please refer to my paper"An Improved LSTM Model for Behavior Recognition of Intelligent Vehicles" for vehicle behavior recognition.