Selected work

Representative Papers

Supporting Perspective Acquisition and Opinion Formation on Societal Issues Through AI-Generated Japanese Rap Battle Debates

Oct 01, 2026

This study addresses the high time costs and accessibility barriers of traditional debates, which limit public exposure to diverse perspectives. We propose an AI-driven framework that generates Japanese rap-battle-style debates by leveraging a large language model (GPT-5.4-mini) and speech synthesis to automatically produce rhyming adversarial scripts on social issues, presented through a web interface. Experimental results demonstrate that this approach doubles the number of viewpoints identified by users, with 52.5% of participants reconsidering or revising their prior stances. By innovatively introducing rap battles into educational contexts, this work validates the effectiveness of AI-generated adversarial content in facilitating low-cost, highly engaging perspective acquisition.

0 citationsRead paper

QC-GAN: A Parameter-Efficient Quaternion Conformer GAN for High-Fidelity Speech Enhancement

Jun 16, 2026

This work addresses the challenge of balancing model efficiency and performance in high-fidelity speech enhancement by proposing QC-GAN, a novel framework that integrates quaternion representations with the Conformer architecture for the first time. By leveraging Hamiltonian products to jointly model magnitude and phase in a structured manner, QC-GAN preserves their intrinsic correlation while substantially reducing parameter count. The approach further incorporates the MetricGAN training strategy and a metric learning-based discriminator to optimize perceptual quality. On the VoiceBank+DEMAND dataset, the model achieves a PESQ score of 3.48 with only 0.89 million parameters, and even a compact 35K-parameter variant attains 3.23—significantly outperforming conventional methods. Strong generalization capability is also demonstrated on the DNS-Challenge 3 benchmark.

0 citationsRead paper

Quaternion Self-Attention with Shared Scores

May 24, 2026

Existing quaternion self-attention mechanisms compute attention scores independently for each quaternion component, resulting in high computational overhead and inconsistent attention distributions. This work proposes a shared-score quaternion self-attention mechanism that generates a single real-valued attention score via quaternion inner product and shares the resulting attention distribution across all components. Theoretical analysis reveals that when queries and keys are pre-mixed through quaternion linear projections, component-wise independent scoring and shared scoring operate within the same interaction subspace, with the former merely constituting a reparameterization of the latter without enhancing representational capacity. Experiments demonstrate that the proposed method reduces GPU and CPU inference time by 44.3% and 58.1%, respectively, on speech enhancement tasks while maintaining performance, and consistently yields advantages across vision and natural language processing benchmarks.

0 citationsRead paper

Recent publications

Latest Papers

Supporting Perspective Acquisition and Opinion Formation on Societal Issues Through AI-Generated Japanese Rap Battle Debates

Oct 01, 2026

This study addresses the high time costs and accessibility barriers of traditional debates, which limit public exposure to diverse perspectives. We propose an AI-driven framework that generates Japanese rap-battle-style debates by leveraging a large language model (GPT-5.4-mini) and speech synthesis to automatically produce rhyming adversarial scripts on social issues, presented through a web interface. Experimental results demonstrate that this approach doubles the number of viewpoints identified by users, with 52.5% of participants reconsidering or revising their prior stances. By innovatively introducing rap battles into educational contexts, this work validates the effectiveness of AI-generated adversarial content in facilitating low-cost, highly engaging perspective acquisition.

0 citationsRead paper

QC-GAN: A Parameter-Efficient Quaternion Conformer GAN for High-Fidelity Speech Enhancement

Jun 16, 2026

This work addresses the challenge of balancing model efficiency and performance in high-fidelity speech enhancement by proposing QC-GAN, a novel framework that integrates quaternion representations with the Conformer architecture for the first time. By leveraging Hamiltonian products to jointly model magnitude and phase in a structured manner, QC-GAN preserves their intrinsic correlation while substantially reducing parameter count. The approach further incorporates the MetricGAN training strategy and a metric learning-based discriminator to optimize perceptual quality. On the VoiceBank+DEMAND dataset, the model achieves a PESQ score of 3.48 with only 0.89 million parameters, and even a compact 35K-parameter variant attains 3.23—significantly outperforming conventional methods. Strong generalization capability is also demonstrated on the DNS-Challenge 3 benchmark.

0 citationsRead paper

Quaternion Self-Attention with Shared Scores

May 24, 2026

Existing quaternion self-attention mechanisms compute attention scores independently for each quaternion component, resulting in high computational overhead and inconsistent attention distributions. This work proposes a shared-score quaternion self-attention mechanism that generates a single real-valued attention score via quaternion inner product and shares the resulting attention distribution across all components. Theoretical analysis reveals that when queries and keys are pre-mixed through quaternion linear projections, component-wise independent scoring and shared scoring operate within the same interaction subspace, with the former merely constituting a reparameterization of the latter without enhancing representational capacity. Experiments demonstrate that the proposed method reduces GPU and CPU inference time by 44.3% and 58.1%, respectively, on speech enhancement tasks while maintaining performance, and consistently yields advantages across vision and natural language processing benchmarks.

0 citationsRead paper