Cognitive/mental task classification using single/limited channel(s) electroencephalogram (EEG) signals in real-time play an important role in designing portable brain-computer interface (BCI) and neurofeedback (NFB) systems. However, real-time recorded EEG signals are often contaminated with ocular artifacts (OAs) and muscle artifacts (MAs), which deteriorate the hand-crafted features extracted from EEG signal, resulting inadequate classification of cognitive tasks. Therefore, we investigate the use of deep learning techniques which do not require manual feature extraction or artifact suppression. In this paper, we propose a shallow one-dimensional convolutional neural network (1D-CNN) architecture for cognitive task classification. The robustness of the proposed architecture is evaluated using artifact-free and artifact-contaminated EEG signals taken from two publicly available databases (i.e, Keirn and Aunon (K) database and EEGMAT (E) database) and in-house (R) database recorded using single-channel device in performing not only cognitive/non-cognitive binary task classification but also cognitive/cognitive multi-tasks classification. Evaluation results demonstrate that the proposed architecture achieves the highest subject-independent classification accuracy of 99.70% and 100.00% for multi-class classification and pair-wise classification respectively in database K. Further, subject-independent classification accuracies of 99.00% and 98.00% are achieved in databases E and R respectively. Comparative performance analysis demonstrates that the proposed architecture outperforms existing approaches not only in terms of classification accuracy but also in robustness against artifacts. © 2022 Elsevier Ltd