Phuong-Nam Tran, Ph.D.c

tpnam0901@gmail.com

tpnam0901@khu.ac.kr

AI Researcher

Short Bio: My name is Tran Phuong Nam (Trần Phương Nam - Vietnamese). I received a B.Sc degree in Artificial Intelligence from FPT University, Vietnam, in 2023.

From 2021 to 2023, I worked as a Junior AI Engineer at K&G company in Vietnam. During this time, I also worked as an AI researcher at startup Pythera AI, focusing on projects related to Natural Language Processing (NLP) and Computer Vision (CV). After that, I worked as a researcher at FPT University in Ho Chi Minh City, Vietnam, under the supervision of Dr. Duc Ngoc Minh Dang. Currently, I am pursuing my Ph.D. studies at Kyung Hee University in the Republic of Korea and being a member of the Networking Intelligence Laboratory led by Prof. Choong Seon Hong.

Research Interests: Medical image segmentation (Primary research), Federated Learning (Primary research), Image segmentation, Multi-Modal and their applications in Biometrics and Audio & Speech Processing, and Computer Vision.

📰News📰

Feb 4, 2025	🗞️ Paper "MemoCMT Multimodal Emotion Recognition Using Cross-Modal Transformer-Based Feature Fusion" was accepted by the Scientific Reports.
Feb 25, 2024	🗞️ 01 paper was accepted by the Expert Systems With Applications Journal (Q1).
Feb 25, 2024	🗞️🗞️🗞️🗞️ 04 papers were accepted at The 15th International Conference on ICT Convergence, Jeju, Korea, Oct 16 -18, 2024.
Dec 29, 2023	🗞️ 01 paper was accepted at The EAI INISCOM 2024 - 10th EAI International Conference on Industrial Networks and Intelligent Systems
Dec 23, 2023	🗞️ 01 paper was accepted at The 38th International Conference on Information Networking (ICOIN 2024)

📑Selected publications📑

Multi-modal Speech Emotion Recognition: Improving Accuracy Through Fusion of VGGish and BERT Features with Multi-head Attention

Phuong-Nam Tran, Thuy-Duong Thi Vu, Nhat Truong Pham, Duc Ngoc Minh Dang, and Anh-Khoa Tran

In Industrial Networks and Intelligent Systems, 2023

Abs HTML Code

Recent research has shown that multi-modal learning is a successful method for enhancing classification performance by mixing several forms of input, notably in speech-emotion recognition (SER) tasks. However, the difference between the modalities may affect SER performance. To overcome this problem, a novel approach for multi-modal SER called 3M-SER is proposed in this paper. The 3M-SER leverages multi-head attention to fuse information from multiple feature embeddings, including audio and text features. The 3M-SER approach is based on the SERVER approach but includes an additional fusion module that improves the integration of text and audio features, leading to improved classification performance. To further enhance the correlation between the modalities, a LayerNorm is applied to audio features prior to fusion. Our approach achieved an unweighted accuracy (UA) and weighted accuracy (WA) of 79.96% and 80.66%, respectively, on the IEMOCAP benchmark dataset. This indicates that the proposed approach is better than SERVER and recent methods with similar approaches. In addition, it highlights the effectiveness of incorporating an extra fusion module in multi-modal learning.
Comparative analysis of multi-loss functions for enhanced multi-modal speech emotion recognition

Phuong-Nam Tran, Thuy-Duong Thi Vu, Nhat Truong Pham, Hanh Dang-Ngoc, and Duc Ngoc Minh Dang

In 2023 14th International Conference on Information and Communication Technology Convergence (ICTC), 2023

Abs HTML Code

In recent years, multi-modal analysis has gained significant prominence across domains such as audio/speech processing, natural language processing, and affective computing, with a particular focus on speech emotion recognition (SER). The integration of data from diverse sources, encompassing text, audio, and images, in conjunction with classifier algorithms has led to the realization of enhanced performance in SER tasks. Traditionally, the cross-entropy loss function has been employed for the classification problem. However, it is challenging to discriminate the feature representations among classes for multi-modal classification tasks. In this study, we focus on the impact of the loss functions on multi-modal SER rather than designing the model architecture. Mainly, we evaluate the performance of multi-modal SER with different loss functions, such as cross-entropy loss, center loss, contrastive-center loss, and their combinations. Based on extensive comparative analysis, it is proven that the combination of cross-entropy loss and contrastive-center loss achieves the best performance for multi-modal SER. This combination reaches the highest accuracy of 80.27% and the highest balanced accuracy of 81.44% on the IEMOCAP dataset.