Search

Many sexually mature women experience premenstrual syndrome (PMS) or premenstrual dysphoric mood disorder (PMDD). Current approaches for managing PMS and PMDD rely on daily mental condition recording, which many discontinue due to its impracticality. Hence, there's a critical need for a simple, objective method to monitor mental symptoms. One of the principal symptoms of PMDD is a dysfunction in emotional regulation, which has been demonstrated through brain-function imaging measurements to involve hyperactivity in the amygdala and a decrease in functionality in the prefrontal cortex (PFC). However, most research has been focused on PMDD, leaving a gap in understanding of PMS. Near-infrared spectroscopy (NIRS) measures brain activity by spectroscopically determining the amount of hemoglobin in the blood vessels. This study aimed to characterize the emotional regulation function in PMS. We measured brain activity in the PFC region using NIRS when participants were presented with emotion-inducing pictures. Furthermore, moods highly associated with emotions were assessed through questionnaires. Forty-six participants were categorized into non-PMS, PMS, and PMDD groups based on the gynecologist's diagnosis. POMS2 scores revealed higher negative mood and lower positive mood in the follicular phase for the PMS group, while the PMDD group exhibited heightened negative mood during the luteal phase. NIRS results showed reduced emotional expression in the PMS group during both phases, while no significant differences were observed in the PMDD group compared to non-PMS. It was found that there are differences in the distribution of mood during the luteal and follicular phase and in cerebral blood flow responses to emotional stimuli between PMS and PMDD. These findings suggest the potential for providing individuals with awareness of PMS or PMDD through scores on the POMS2 and NIRS measurements., Comment: 5 pages, 4 figures, 1 table, 2024 IEEE International Conference on Robotics and Automation (ICRA 2024)

Deep neural networks have achieved remarkable success in a variety of computer vision applications. However, there is a problem of degrading accuracy when the data distribution shifts between training and testing. As a solution of this problem, Test-time Adaptation~(TTA) has been well studied because of its practicality. Although TTA methods increase accuracy under distribution shift by updating the model at test time, using high-uncertainty predictions is known to degrade accuracy. Since the input image is the root of the distribution shift, we incorporate a new perspective on enhancing the input image into TTA methods to reduce the prediction's uncertainty. We hypothesize that enhancing the input image reduces prediction's uncertainty and increase the accuracy of TTA methods. On the basis of our hypothesis, we propose a novel method: Test-time Enhancer and Classifier Adaptation~(TECA). In TECA, the classification model is combined with the image enhancement model that transforms input images into recognition-friendly ones, and these models are updated by existing TTA methods. Furthermore, we found that the prediction from the enhanced image does not always have lower uncertainty than the prediction from the original image. Thus, we propose logit switching, which compares the uncertainty measure of these predictions and outputs the lower one. In our experiments, we evaluate TECA with various TTA methods and show that TECA reduces prediction's uncertainty and increases accuracy of TTA methods despite having no hyperparameters and little parameter overhead., Comment: Accepted to IJCNN2024

This paper addresses the tradeoff between standard accuracy on clean examples and robustness against adversarial examples in deep neural networks (DNNs). Although adversarial training (AT) improves robustness, it degrades the standard accuracy, thus yielding the tradeoff. To mitigate this tradeoff, we propose a novel AT method called ARREST, which comprises three components: (i) adversarial finetuning (AFT), (ii) representation-guided knowledge distillation (RGKD), and (iii) noisy replay (NR). AFT trains a DNN on adversarial examples by initializing its parameters with a DNN that is standardly pretrained on clean examples. RGKD and NR respectively entail a regularization term and an algorithm to preserve latent representations of clean examples during AFT. RGKD penalizes the distance between the representations of the standardly pretrained and AFT DNNs. NR switches input adversarial examples to nonadversarial ones when the representation changes significantly during AFT. By combining these components, ARREST achieves both high standard accuracy and robustness. Experimental results demonstrate that ARREST mitigates the tradeoff more effectively than previous AT-based methods do., Comment: Accepted by International Conference on Computer Vision (ICCV) 2023

This paper proposes a novel automatic speech recognition (ASR) system that can transcribe individual speaker's speech while identifying whether they are target or non-target speakers from multi-talker overlapped speech. Target-speaker ASR systems are a promising way to only transcribe a target speaker's speech by enrolling the target speaker's information. However, in conversational ASR applications, transcribing both the target speaker's speech and non-target speakers' ones is often required to understand interactive information. To naturally consider both target and non-target speakers in a single ASR model, our idea is to extend autoregressive modeling-based multi-talker ASR systems to utilize the enrollment speech of the target speaker. Our proposed ASR is performed by recursively generating both textual tokens and tokens that represent target or non-target speakers. Our experiments demonstrate the effectiveness of our proposed method., Comment: Accepted at Interspeech 2023

This paper investigates the effectiveness and implementation of modality-specific large-scale pre-trained encoders for multimodal sentiment analysis~(MSA). Although the effectiveness of pre-trained encoders in various fields has been reported, conventional MSA methods employ them for only linguistic modality, and their application has not been investigated. This paper compares the features yielded by large-scale pre-trained encoders with conventional heuristic features. One each of the largest pre-trained encoders publicly available for each modality are used; CLIP-ViT, WavLM, and BERT for visual, acoustic, and linguistic modalities, respectively. Experiments on two datasets reveal that methods with domain-specific pre-trained encoders attain better performance than those with conventional features in both unimodal and multimodal scenarios. We also find it better to use the outputs of the intermediate layers of the encoders than those of the output layer. The codes are available at https://github.com/ando-hub/MSA_Pretrain., Comment: Accepted to SLT 2022

In this paper, we investigate the semi-supervised joint training of text to speech (TTS) and automatic speech recognition (ASR), where a small amount of paired data and a large amount of unpaired text data are available. Conventional studies form a cycle called the TTS-ASR pipeline, where the multispeaker TTS model synthesizes speech from text with a reference speech and the ASR model reconstructs the text from the synthesized speech, after which both models are trained with a cycle-consistency loss. However, the synthesized speech does not reflect the speaker characteristics of the reference speech and the synthesized speech becomes overly easy for the ASR model to recognize after training. This not only decreases the TTS model quality but also limits the ASR model improvement. To solve this problem, we propose improving the cycleconsistency-based training with a speaker consistency loss and step-wise optimization. The speaker consistency loss brings the speaker characteristics of the synthesized speech closer to that of the reference speech. In the step-wise optimization, we first freeze the parameter of the TTS model before both models are trained to avoid over-adaptation of the TTS model to the ASR model. Experimental results demonstrate the efficacy of the proposed method., Comment: Accepted to INTERSPEECH 2022

Execution of quantum algorithms requires a quantum computer architecture with a dedicated quantum instruction set that is capable of supporting translation of workloads into actual quantum operations acting on the qubits. State-of-the-art qubit control setups typically utilize general purpose test instruments such as arbitrary waveform generators (AWGs) to generate a limited set of waveforms or pulses. These waveforms are precomputed and stored prior to execution, and then used to produce control pulses during execution. Besides their prohibitive cost and limited scalability, such instruments suffer from poor programmability due to the absence of an instruction set architecture (ISA). Limited memory for pulse storage ultimately determines the total number of supported quantum operations. In this work, we present a scalable qubit control system that enables efficient qubit control using a flexible ISA to drive a direct digital synthesis (DDS) pipeline producing nanosecond-accurate qubit control signals dynamically. The designed qubit controller provides a higher density of control channels, a scalable design, better programmability, and lower cost compared to state-of-the-art systems. In this work, we discuss the new qubit controller's capabilities, its architecture and instruction set, and present experimental results for coherent qubit control., Comment: 10 pages