Srinivasa Rao Chalamala

The Large Language Models (LLMs) exhibit remarkable ability to generate fluent content across a wide spectrum of user queries. However, this capability has raised concerns regarding misinformation and personal information leakage. In this paper, we present our methods for the SemEval2024 Task8, aiming to detect machine-generated text across various domains in both mono-lingual and multi-lingual contexts. Our study comprehensively analyzes various methods to detect machine-generated text, including statistical, neural, and pre-trained model approaches. We also detail our experimental setup and perform a in-depth error analysis to evaluate the effectiveness of these methods. Our methods obtain an accuracy of 86.9\% on the test set of subtask-A mono and 83.7\% for subtask-B. Furthermore, we also highlight the challenges and essential factors for consideration in future studies.

Ashok Urlana, Aditya Saibewar, Bala Mallikarjunarao Garlapati, Charaka Vinayak Kumar, Ajeet Singh, Srinivasa Rao Chalamala. Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024). 2024.

A Fuzzy Approach to Mute Sensitive Information in Noisy Audio Conversations

Digital watermarking is a process to provide authenticity by hiding a data into an image or audio or document. Hiding of data in an image can be done in frequency domain. Since frequency domain based techniques are more robust against signal processing and geometric attacks than time domain based techniques and watermark can also be extracted with(non blind) and without(blind) original cover image. In this paper we proposed a blind image watermarking technique which embeds watermark into image in frequency domain using discrete wavelet transform, singular value decomposition and torus automorphism techniques. This technique extracts watermark without cover image and also proved that this method is robust against different signal and non signal processing attacks.

Digital watermarking is a process to provide authenticity by hiding a data into an image or audio or document. Hiding of data in an image can be done in frequency domain. Since frequency domain based techniques are more robust against signal processing and geometric attacks than time domain techniques. One such frequency domain technique is developed based on combination of discrete wavelet transform, singular value decomposition and torus automorphism and is presented in this paper. This results prove that this method robust against different attacks and has good PSNR ratio.