Lochan Basyal

Lochan Basyal

Machine Learning Engineer | Data Scientist | Generative-AI | Computer Vision

Talks and presentations

Text Summarization Using Large Language Models (LLMs)

October 29, 2023

Talk, KaggleX BIPOC Mentorship Program 2023, Cohort - 3, New Jersey, USA

Text summarization is a critical Natural Language Processing (NLP) task with applications ranging from information retrieval to content generation. Leveraging Large Language Models (LLMs) has shown remarkable promise in enhancing summarization techniques. This paper embarks on an exploration of text summarization with a diverse set of LLMs, including MPT-7b-instruct, falcon-7b-instruct, and OpenAI ChatGPT text-davinci-003 models. The experiment was performed with different hyperparameters and evaluated the generated summaries using widely accepted metrics such as the Bilingual Evaluation Understudy (BLEU) Score, Recall-Oriented Understudy for Gisting Evaluation (ROUGE) Score, and Bidirectional Encoder Representations from Transformers (BERT) Score. According to the experiment, text-davinci-003 outperformed the others. This investigation involved two distinct datasets: CNN Daily Mail and XSum. Its primary objective was to provide a comprehensive understanding of the performance of Large Language Models (LLMs) when applied to different datasets. The assessment of these models’ effectiveness contributes valuable insights to researchers and practitioners within the NLP domain. This work serves as a resource for those interested in harnessing the potential of LLMs for text summarization and lays the foundation for the development of advanced Generative AI applications aimed at addressing a wide spectrum of business challenges.

Histopathologic Cancer Detection

December 21, 2022

Talk, Stevens Institute of Technology, Hoboken, New Jersey, USA

Early diagnosis of the cancer cells is necessary for making an effective treatment plan and the health safety of a patient. Nowadays, doctors usually use a histological grade that pathologists determine by performing a semi-quantitative analysis of the histopathological and cytological features of hematoxylin-eosin (HE) stained histopathological images. This research contributes a potential classification model for cancer prognosis to efficiently utilize the valuable information underlying the HE-stained histopathological images. This work uses the PatchCamelyon benchmark datasets and trains them in a multi- layer perceptron and convolution model to observe the model’s performance in terms of precision, Recall, F1 Score, Accuracy, and AUC Score. The evaluation result shows that the baseline convolution model outperforms the baseline MLP model. Also, this paper introduced ResNet50 and InceptionNet models with data augmentation where ResNet50 able to beat the state-of- the-art model. Furthermore, majority vote and concatenation ensemble were evaluated and provides the future direction of using transfer learning, and segmentation to understand the specific features.

Implementation of BCI Technology in Real-time Automation

May 05, 2022

Talk, Stevens Institute of Technology, Hoboken, New Jersey, USA

The field of Brain-Computer Interface (BCI) technology is rapidly advancing, with researchers working towards establishing a direct communication channel between the human brain and computers or external devices. This collaborative effort aims to facilitate communication by exchanging information derived from neural activities in the brain with the computer. A Brain-Computer Interface involves the seamless integration of the brain with mechanical devices, where the brain naturally accepts and controls these devices as an intrinsic part of its body representation. This innovative technology extracts electrophysiology signals from specific brain components and processes them to generate control signals for computers, robotic machines, or communication devices. The evolving landscape of BCI holds promise for revolutionizing the way humans interact with technology, opening up new avenues for enhanced communication and control in various applications.

Smart Home Application: Lochan Control System

October 05, 2018

Talk, IISF Young Scientists' Conference 2018, Lucknow, India

Smart Home Application is an emerging concept for uplifting society as well as a country in the direction of advanced technology. Lochan Control System is based on Home Automation Technique through which we can control all the electrical appliances that are connected in our home, offices as well as a company through an android application. Today as we see must of the people uses a smartphone and which is based on an Android platform. If we adopt this proposed system on every citizen home then the concept of manual switching has been transformed into an Automatic Switching and people around there live their life more easily. This approach helps a disabled people as well as normal people to perform their respective task. Lochan Control System is based on Microcontroller Atmega328 with interfacing Bluetooth Module HC-05 as a serial communication with an android application. There is no any operating cost for this project but it has some amounts nearly 1500 INR for Initial installation. This project was successfully implemented with features of low manufacturing cost, compact size, high system response, voice controlled, less power consumption and no operating cost.