Publications

2025

1. COSYNE

   The posterior parietal cortex mediates serial dependence during visuospatial attention

   Raj V Jain, and Devarajan Sridharan

   In Computational and Systems Neuroscience, Mar 2025

   Bib

   @inproceedings{ppc_sd_cosyne,
     author = {Jain, Raj V and Sridharan, Devarajan},
     title = {The posterior parietal cortex mediates serial dependence during visuospatial attention},
     year = {2025},
     month = mar,
     booktitle = {Computational and Systems Neuroscience}
   }

2024

1. CCN

   A Causal Role for the Posterior Parietal Cortex in Mediating Serial Dependence during Visuospatial Attention

   Raj V Jain, Ankita Sengupta, and Devarajan Sridharan

   In Cognitive Computational Neuroscience, Aug 2024

   Abs Bib PDF

   Events in the recent past – even those that are no longer relevant – may be tracked implicitly by the brain and influence our decisions, a phenomenon known as "serial dependence." The precise role of the posterior parietal cortex (PPC) in mediating serial dependence remains actively researched. Here, we evaluate serial dependencies in behavior with a visuospatial attention task (n=26 participants, n=39000 trials). Training a long short-term memory network (LSTM) to predict participants’ trial-wise responses from task variable history, we identify robust serial dependency effects in reaction times. 40-Hz transcranial alternating current stimulation (tACS) over PPC significantly mitigates the magnitude of these serial dependence effects. Expected gradients-based feature attribution traced tACS effects to a reduced impact of selection history associated with attentional cueing. The results reveal a causal role for the PPC in mediating serial dependence in "experience-driven attention”, with critical implications for understanding attentional mechanisms in the human brain.

   @inproceedings{ppc_sd_ccn,
     author = {Jain, Raj V and Sengupta, Ankita and Sridharan, Devarajan},
     title = {A Causal Role for the Posterior Parietal Cortex in Mediating Serial Dependence during Visuospatial Attention},
     year = {2024},
     month = aug,
     booktitle = {Cognitive Computational Neuroscience},
   }

2023

1. ACCS

   Discovering Serial Dependence with SequenceAware Neural Networks

   Raj V. Jain, and Devarajan Sridharan

   In Annual Conference of Cognitive Science, Dec 2023

   Abs Bib PDF

   Serial dependence is the phenomenon where past experiences influence our current behavior, even when these past experiences are objectively irrelevant to the task at hand [1]. We have chosen three experiments performed in our lab, which invoke different cognitive processes (see next section), to quantify serial dependence and understand how cognitive factors influence it. Previously, serial dependence has been studied with models that incorporate strong, and often simplifying, assumptions on the decision-making process [2], [3]. Here, we use powerful sequence-aware neural networks (SANNs) that enable extracting sensitive metrics of serial dependence directly from behavioral data. SANNs fall into one of various types, such as conventional recurrent neural networks (RNNs), long short-term memory networks (LSTMs) [4] and position-encoded Transformers [5]. Here, we employed LSTMs for serial dependence detection.

   @inproceedings{sd_accs,
     author = {Jain, Raj V. and Sridharan, Devarajan},
     title = {Discovering Serial Dependence with SequenceAware Neural Networks},
     year = {2023},
     month = dec,
     booktitle = {Annual Conference of Cognitive Science},
   }

2021

1. FICTA

   HMM Classifier Object Recognizing System in Brain–Computer Interface

   H. S. Anupama, Raj V. Jain, Revannur Venkatesh, and 2 more authors

   In Evolution in Computational Intelligence, Dec 2021

   Abs Bib

   Machine learning (ML) is the field that adds intelligence to devices providing them with capabilities to process and identify patterns in data just like human beings do. Programming devices in this manner can help in identifying those patterns which human beings often cannot. Machine learning is based on modelling data mathematically. ML has been gaining a lot of attention in the last few decades, especially in fields of interdisciplinary research. Brain–Computer Interface (BCI) is an area where Machine Learning Technology is been rapidly using. Also, Machine Learning techniques have to be used so that one can get a better result and more efficiency. Information Transfer Rate is the best way to measure the performance of the signals. The current research is mainly focused on achieving the systems with higher ITR. The focus of the proposed system is to get better and high Information Transfer Rate by merging two different approaches. The approach used in this work is (SSVEP), Visually Evoked Potential and (SSAEP) Auditory Evoked Potential by using Hidden Markova Model (HMM). The system which is to be developed checks whether the existing system has such facility if it has, does it provides accuracy which is of a higher rate and can put it in the real-world applications.

   @inproceedings{hmm_cmovabci,
     author = {Anupama, H. S. and Jain, Raj V. and Venkatesh, Revannur and Cauvery, N. K. and Lingaraju, G. M.},
     editor = {Bhateja, Vikrant and Peng, Sheng-Lung and Satapathy, Suresh Chandra and Zhang, Yu-Dong},
     title = {HMM Classifier Object Recognizing System in Brain--Computer Interface},
     booktitle = {Evolution in Computational Intelligence},
     year = {2021},
     publisher = {Springer Singapore},
     address = {Singapore},
     pages = {287-294},
     isbn = {978-981-15-5788-0}
   }

2018

1. ICACNI

   Implementing and Analyzing Different Feature Extraction Techniques Using EEG-Based BCI

   H. S. Anupama, Raj V. Jain, Revanur Venkatesh, and 3 more authors

   In Recent Findings in Intelligent Computing Techniques, Dec 2018

   Abs Bib

   Brain computer interface (BCI) is a method of communication between the brain and computer or machines, which use the neural activity of the brain. This neural activity communication does not occur using the peripheral nervous system and muscles, as is the usual case in human beings, but through any other mechanism. This paper focuses on different types of feature extraction techniques to explore a new kind of BCI paradigm and validate whether it can give a better ITR as compared to the existing paradigms.

   @inproceedings{features_cmovabci,
     author = {Anupama, H. S. and Jain, Raj V. and Venkatesh, Revanur and Mahadevan, Rupa and Cauvery, N. K. and Lingaraju, G. M.},
     editor = {Sa, Pankaj Kumar and Bakshi, Sambit and Hatzilygeroudis, Ioannis K. and Sahoo, Manmath Narayan},
     title = {Implementing and Analyzing Different Feature Extraction Techniques Using EEG-Based BCI},
     booktitle = {Recent Findings in Intelligent Computing Techniques},
     year = {2018},
     publisher = {Springer Singapore},
     address = {Singapore},
     pages = {377-386},
     isbn = {978-981-10-8636-6}
   }

2017

1. IJAER

   Implementing and analyzing different Machine Learning Algorithms using EEG based BCI

   H. S. Anupama, Raj V. Jain, Revannur Venkatesh, and 2 more authors

   International Journal of Applied Engineering Research, Dec 2017

   Abs Bib PDF

   Communication that exists between Human Brain and Computer is termed as Brain Computer Interface (BCI). This uses neuronal activity of the brain. Information is passed from one part of the body to another through the neurons present in the human brain. This paper focuses on different types of classification algorithm to explore a new kind of BCI paradigm. This also focuses on the Information Transfer Rate of the existing paradigms and also compares our paradigm with the existing paradigms.

   @article{ml_cmovabci,
     author = {Anupama, H. S. and Jain, Raj V. and Venkatesh, Revannur and Cauvery, N. K. and Lingaraju, G. M.},
     title = {Implementing and analyzing different Machine Learning Algorithms using EEG based BCI},
     journal = {International Journal of Applied Engineering Research},
     volume = {12},
     number = {8},
     pages = {1736-1741},
     year = {2017},
     issn = {0973-4562},
   }