Zheng-Hua Tan
YOU?
Author Swipe
View article: Detecting and Defending Against Adversarial Attacks on Automatic Speech Recognition via Diffusion Models
Detecting and Defending Against Adversarial Attacks on Automatic Speech Recognition via Diffusion Models Open
Automatic speech recognition (ASR) systems are known to be vulnerable to adversarial attacks. This paper addresses detection and defence against targeted white-box attacks on speech signals for ASR systems. While existing work has utilised…
View article: Soil Moisture Retrieval with High Spatial-temporal Resolution by Fusion of CYGNSS and SAR Data
Soil Moisture Retrieval with High Spatial-temporal Resolution by Fusion of CYGNSS and SAR Data Open
As the probability and intensity of global drought events continue to increase, soil moisture is an important basis for drought monitoring and drought assessment, and the need to accurately obtain soil moisture distribution information wit…
View article: Bridging the Evaluation Gap: Leveraging Large Language Models for Topic Model Evaluation
Bridging the Evaluation Gap: Leveraging Large Language Models for Topic Model Evaluation Open
This study presents a framework for automated evaluation of dynamically evolving topic taxonomies in scientific literature using Large Language Models (LLMs). In digital library systems, topic modeling plays a crucial role in efficiently o…
View article: Error Tolerance in Automatic Short Answer Grading with Large Language Models: The Case of Handwriting Recognition Errors
Error Tolerance in Automatic Short Answer Grading with Large Language Models: The Case of Handwriting Recognition Errors Open
Assessing students' answers is a labor-intensive task that could significantly burden educators. Technological advances have led to the development of automatic scoring systems. However, handwriting recognition errors significantly impacts…
View article: Masked Autoencoders with Multi-Window Local-Global Attention Are Better Audio Learners
Masked Autoencoders with Multi-Window Local-Global Attention Are Better Audio Learners Open
In this work, we propose a Multi-Window Masked Autoencoder (MW-MAE) fitted with a novel Multi-Window Multi-Head Attention (MW-MHA) module that facilitates the modelling of local-global interactions in every decoder transformer block throug…
View article: On the Contributions of the Northeast Cold Vortex to the Formation and Evolution of Backflow Blizzard
On the Contributions of the Northeast Cold Vortex to the Formation and Evolution of Backflow Blizzard Open
Blizzard is a severe weather-related disaster with significant environmental, economic, and social impacts. What is worse, the blizzard is increasingly frequent along with climate change. To enhance resilience, it is important to accuratel…