Research Scholar
- Image data set creation and annotation
- Deep learning Model Development & training on supercomputer for object detection, object classification and 3D object detection
- Quantization of Deep learning models to execute on resource constraint hardwares like Raspberry pi, Jetson nano and similar
- Hardware development and deployment of models on edge devices like raspberry pi
- Accelerating the model performance using Intel Neural compute stick 2 with edge devices.
- 3D Object detection using depth camera like ZED Mini
Patent:
Patent Title: “System and Method for Aiding Visually Impaired person”.
Inventor/Applicant: K. Manjari, M. Verma and G. Singal
Publication Date: June 26, 2020
Application Number: 202011024622
URL: https://ipindiaservices.gov.in/PatentSearch/PatentSearch/ViewApplicationStatus
Conferences:
- Manjari, Kanak, Madhushi Verma, and Gaurav Singal. "A Travel Aid for Visually Impaired: R-Cane." International Conference on Smart City and Informatization. Springer, Singapore, 2019.
- Manjari, Kanak, Madhushi Verma, and Gaurav Singal. "CREATION: Computational ConstRained Travel Aid for Object Detection in Outdoor eNvironment." 2019 15th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS). IEEE, 2019.
Book Chapters:
- Manjari, Kanak, Madhushi Verma, and Gaurav Singal. "A smart hand for VI: Resource-constrained assistive technology for visually impaired." IoT-Based Data Analytics for the Healthcare Industry. Academic Press, 2021. 225-235.
- Manjari, Kanak, et al. "AnimNet: An Animal Classification Network using Deep Learning." Machine Learning Algorithms and Applications (2021): 247-265.
Journals:
- Manjari, Kanak, Madhushi Verma, and Gaurav Singal. "A survey on assistive technology for visually impaired." Internet of Things 11 (2020): 100188.
- Manjari, Kanak, Madhushi Verma, and Gaurav Singal. "Catalysing Assistive Solutions by Deploying Light-Weight Deep Learning Model on Edge Devices." Journal of Experimental and Theoretical Artificial Intelligence, Taylor and Francis, (2021) (ACCEPTED)
- Manjari, Kanak, Madhushi Verma, and Gaurav Singal. "QAOVDetect: A Novel Syllogistic Model with Quantized and Anchor Optimized Approach to Assist Visually Impaired for Animal Detection using 3D Vision." Cognitive Computation, Springer, (2022) (ACCEPTED)
- Manjari, Kanak, Madhushi Verma, and Gaurav Singal. "QEST: Quantized and Efficient Scene Text Detector using Deep Learning." Transactions on Asian and Low-Resource Language Information Processing, ACM (2021) (REVISION)