A5023410809- Multimodal Machine Learning Applications
- Human Pose and Action Recognition
- Advanced Image and Video Retrieval Techniques
- Domain Adaptation and Few-Shot Learning
- Low-power high-performance VLSI design
- Topic Modeling
- Advancements in Semiconductor Devices and Circuit Design
- Analog and Mixed-Signal Circuit Design
University of Illinois Urbana-Champaign
2021
Birla Institute of Technology and Science, Pilani
2015
Fact-based Visual Question Answering (FVQA), a challenging variant of VQA, requires QA-system to include facts from diverse knowledge graph (KG) in its reasoning process produce an answer. Large KGs, especially common-sense are known be incomplete, i.e., not all non-existent always incorrect. Therefore, being able reason over incomplete KGs for QA is critical requirement real-world applications that has been addressed extensively the literature. We develop novel architecture allows us...
Kiran Ramnath, Leda Sari, Mark Hasegawa-Johnson, Chang Yoo. Proceedings of the 2021 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies. 2021.
In this paper, the effects of applying forward body bias technique to an operational amplifier with transistors operating in subthreshold regime are reported. A previously established is chosen as a reference and realized Cadence Virtuoso. The then applied different circuit, studied. Consequently, intelligent method for developed, through identification critical calculation optimum each transistor. Through application proposed methodology given gain increases by 42 percent integrated input...