Priyadarshini Panda

Priya's research interests lie in Neuromorphic Computing: spanning energy-efficient design methodologies for deep learning networks, novel supervised/unsupervised learning algorithms for spiking neural networks and developing neural architectures for new computing scenarios (such as lifelong learning, generative models, stochastic networks, adversarial attacks etc.).

​​​Her goal is to empower energy-aware and energy-efficient machine intelligence through algorithm-hardware co-design while being secure to adversarial scenarios and catering to the resource constraints of Internet of Things (IoT) devices.

A comprehensive list of publications is available here.

Selected Publications (Journal):

• Kaushik Roy, Akhilesh Jaiswal, and Priyadarshini Panda. Towards Spike-based Machine Intelligence with Neuromorphic Computing. To appear in Nature (2019). An online tutorial on the paper encompassing the perspectives on neuromorphic computing field is available on "https://www.youtube.com/watch?v=HnxkQvPcdXs".
• Fan Zuo*, Priyadarshini Panda*, Michele Kotiuga, Jiarui Li, Mingu Kang, Claudio Mazzoli, Hua Zhou et al. Habituation based synaptic plasticity and organismic learning in a quantum perovskite. Nature communications 8, no. 1 (2017): 240 (*Equal author contributions).
• Priyadarshini Panda, Swagath Venkataramani, Abronil Sengupta, Anand Raghunathan, and Kaushik Roy. Energy-efficient object detection using semantic decomposition. IEEE Transactions on Very Large Scale Integration (VLSI) Systems, doi:10.1109/TVLSI.2017.2707077, 25(9):2673–2677, Sept 2017.
• Priyadarshini Panda, Indranil Chakraborty, and Kaushik Roy. Discretization based Solutions for Secure Machine Learning against Adversarial Attacks. IEEE Access (2019).
• Abhronil Sengupta, Priyadarshini Panda, Parami Wijesinghe, Yusung Kim, and Kaushik Roy. Magnetic tunnel junction mimics stochastic cortical spiking neurons. Scientific reports (2016): 30039.
• Deboleena Roy, Priyadarshini Panda, and Kaushik Roy. Tree-cnn: A hierarchical deep convolutional neural network for incremental learning. arXiv preprint arXiv:1802.05800, 2018, Accepted in Neural Networks (Elsevier), 2019.
• Chankyu Lee, Priyadarshini Panda, Gopalakrishnan Srinivasan, and Kaushik Roy. Training deep spiking convolutional neural networks with stdp-based unsupervised pre-training followed by supervised finetuning. Frontiers in Neuroscience, 12:435, 2018.
   

Selected Publications (Conference):

• Priyadarshini Panda, Abhronil Sengupta, and Kaushik Roy. Conditional deep learning for energy-efficient and enhanced pattern recognition. In 2016 Design, Automation & Test in Europe Conference & Exhibition (DATE), pp. 475-480. IEEE, 2016.
• Priyadarshini Panda, and Kaushik Roy. Unsupervised regenerative learning of hierarchical features in spiking deep networks for object recognition. In 2016 International Joint Conference on Neural Networks (IJCNN), pp. 299-306. IEEE, 2016.
• Priyadarshini Panda and Kaushik Roy. Implicit generative modeling of random noise during training for adversarial robustness. arXiv preprint arXiv:1807.02188, In ICML 2019 - Workshop on Uncertainty and Robustness in Deep Learning.
• Priyadarshini Panda, Abhronil Sengupta, Syed Shakib Sarwar, Gopalakrishnan Srinivasan, Swagath Venkataramani, Anand Raghunathan, and Kaushik Roy. Cross-layer approximations for neuromorphic computing: From devices to circuits and systems. In 2016 53nd ACM/EDAC/IEEE Design Automation Conference (DAC), pp. 1-6. IEEE, 2016.
• Aayush Ankit, Abhronil Sengupta, Priyadarshini Panda, and Kaushik Roy. Resparc: A reconfigurable and energy-efficient architecture with memristive crossbars for deep spiking neural networks. In Proceedings of the 54th Annual Design Automation Conference 2017, p. 27. ACM, 2017.