15 February 2021 in Neural Computing and Applications
24 May 2019 in Lecture Notes in Electrical Engineering
1 January 2026 in ITEGAM-Journal of Engineering and Technology for Industrial Applications (ITEGAM-JETIA)
The availability of various sensors in smartphones has made it easier to collect human locomotion data. This data is useful for recognizing different locomotion modes namely walking, cycling and traveling. The correct identification of locomotion modes helps in transportation planning, traffic analysis and travel cost estimation. In most cases, these labels are assigned through crowdsourcing or web-based queries. This paper presents a deep learning-based approach for locomotion mode recognition in the presence of noisy labels. The proposed method builds an ensemble model consisting of three sub-models. The different models were created for processing training data with varying degrees of noise contaminations. A conventional deep learning system obtains spatial-temporal information from basic sensor inputs. The model maintains reliable performance within conditions of minimal label noise. A noise adaptive model contains a dynamic loss function which adjusts its operation according to noise intensity levels. The model shows success in dealing with intense noise levels. The proposed approach showed its effectiveness through experiments conducted on both newly obtained datasets and pre-existing available datasets. This paper resolves the problems with noisy labels which occur in locomotion mode recognition tasks. The ensemble model achieves recognition success through its integration of conventional, adaptive and corrective techniques.
25 December 2016 in International Journal of Computer Trends and Technology
1 January 2026 in ITEGAM-Journal of Engineering and Technology for Industrial Applications (ITEGAM-JETIA)
This paper focuses on identifying visual learners in real time using brain signals. The signals come from an EEG, which records brain activity. Usually, EEG data is noisy and needs cleaning before use. Traditional methods use this cleaned data to find learning styles. But cleaning takes time, so the models are not useful for real-time use. This issue is solved by using raw EEG data. Deep learning models are developed that do not need manual feature extraction. These models learn directly from raw data. The main purpose is to check if someone is a visual learner. A visual learner learns best with images and videos. The EEG data was collected from 34 healthy people. These people had no prior knowledge of the content shown. Participants were shown animated educational videos. While watching, EEG signals were recorded. Three deep learning models were used: LSTM, LSTM-CNN and LSTM-FCNN. These models are good at understanding time- based data like EEG. Among them, LSTM-CNN worked the best. It gave the highest accuracy. It had good sensitivity, specificity and F1-score. This method does not need any self-reports. It gives fast results. It works with different lengths of EEG data. So, it is useful in classrooms and other real-world settings. It helps teachers understand the best way students learn. This improves both teaching and learning. This work shows that deep learning with raw EEG is powerful. It quickly and accurately identifies visual learners. This changes the future of detecting learning styles.