Jane Yung-jen Hsu

Sheng-Wei Li, Zi-Xiang Wei, Wei-Jie Chen et al.

Existing zero-shot skeleton-based action recognition methods utilize projection networks to learn a shared latent space of skeleton features and semantic embeddings. The inherent imbalance in action recognition datasets, characterized by variable skeleton sequences yet constant class labels, presents significant challenges for alignment. To address the imbalance, we propose SA-DVAE -- Semantic Alignment via Disentangled Variational Autoencoders, a method that first adopts feature disentanglement to separate skeleton features into two independent parts -- one is semantic-related and another is irrelevant -- to better align skeleton and semantic features. We implement this idea via a pair of modality-specific variational autoencoders coupled with a total correction penalty. We conduct experiments on three benchmark datasets: NTU RGB+D, NTU RGB+D 120 and PKU-MMD, and our experimental results show that SA-DAVE produces improved performance over existing methods. The code is available at https://github.com/pha123661/SA-DVAE.

Ying-Chieh Hsu, Stanley Yung-Chuan Liu, Chao-Jung Huang et al.

This study investigated the use of deep learning to identify multi-level upper airway collapses in obstructive sleep apnea (OSA) patients based on snoring sounds. We fi-ne-tuned ResNet-50 and Audio Spectrogram Transformer (AST) models using snoring recordings from 37 subjects undergoing drug-induced sleep endoscopy (DISE) between 2020 and 2021. Snoring sounds were labeled according to the VOTE (Velum, Orophar-ynx, Tongue Base, Epiglottis) classification, resulting in 259 V, 403 O, 77 T, 13 E, 1016 VO, 46 VT, 140 OT, 39 OE, 30 VOT, and 3150 non-snoring (N) 0.5-second clips. The models were trained for two multi-label classification tasks: identifying obstructions at V, O, T, and E levels, and identifying retropalatal (RP) and retroglossal (RG) obstruc-tions. Results showed AST slightly outperformed ResNet-50, demonstrating good abil-ity to identify V (F1-score: 0.71, MCC: 0.61, AUC: 0.89), O (F1-score: 0.80, MCC: 0.72, AUC: 0.94), and RP obstructions (F1-score: 0.86, MCC: 0.77, AUC: 0.97). However, both models struggled with T, E, and RG classifications due to limited data. Retrospective analysis of a full-night recording showed the potential to profile airway obstruction dynamics. We expect this information, combined with polysomnography and other clinical parameters, can aid clinical triage and treatment planning for OSA patients.

Bo-Ying Su, Shang-Ling Hsu, Kuan-Yin Lai et al.

Extracting temporal relations among events from unstructured text has extensive applications, such as temporal reasoning and question answering. While it is difficult, recent development of Neural-symbolic methods has shown promising results on solving similar tasks. Current temporal relation extraction methods usually suffer from limited expressivity and inconsistent relation inference. For example, in TimeML annotations, the concept of intersection is absent. Additionally, current methods do not guarantee the consistency among the predicted annotations. In this work, we propose SMARTER, a neural semantic parser, to extract temporal information in text effectively. SMARTER parses natural language to an executable logical form representation, based on a custom typed lambda calculus. In the training phase, dynamic programming on denotations (DPD) technique is used to provide weak supervision on logical forms. In the inference phase, SMARTER generates a temporal relation graph by executing the logical form. As a result, our neural semantic parser produces logical forms capturing the temporal information of text precisely. The accurate logical form representations of an event given the context ensure the correctness of the extracted relations.

Yi-Ching 'Janet' Huang, Yu-Ting Cheng, Rung-Huei Liang et al.

Designing future IoT ecosystems requires new approaches and perspectives to understand everyday practices. While researchers recognize the importance of understanding social aspects of everyday objects, limited studies have explored the possibilities of combining data-driven patterns with human interpretations to investigate emergent relationships among objects. This work presents Thing Constellation Visualizer (thingCV), a novel interactive tool for visualizing the social network of objects based on their co-occurrence as computed from a large collection of photos. ThingCV enables perspective-changing design explorations over the network of objects with scalable links. Two exploratory workshops were conducted to investigate how designers navigate and make sense of a network of objects through thingCV. The results of eight participants showed that designers were actively engaged in identifying interesting objects and their associated clusters of related objects. The designers projected social qualities onto the identified objects and their communities. Furthermore, the designers changed their perspectives to revisit familiar contexts and to generate new insights through the exploration process. This work contributes a novel approach to combining data-driven models with designerly interpretations of thing constellation towards More-Than Human-Centred Design of IoT ecosystems.

Shan-Jean Wu, Chih-Yuan Yang, Jane Yung-jen Hsu

Chinese calligraphy is the writing of Chinese characters as an art form performed with brushes so Chinese characters are rich of shapes and details. Recent studies show that Chinese characters can be generated through image-to-image translation for multiple styles using a single model. We propose a novel method of this approach by incorporating Chinese characters' component information into its model. We also propose an improved network to convert characters to their embedding space. Experiments show that the proposed method generates high-quality Chinese calligraphy characters over state-of-the-art methods measured through numerical evaluations and human subject studies.

Yi-Ching Huang, Yu-Ting Cheng, Lin-Lin Chen et al.

Human and AI are increasingly interacting and collaborating to accomplish various complex tasks in the context of diverse application domains (e.g., healthcare, transportation, and creative design). Two dynamic, learning entities (AI and human) have distinct mental model, expertise, and ability; such fundamental difference/mismatch offers opportunities for bringing new perspectives to achieve better results. However, this mismatch can cause unexpected failure and result in serious consequences. While recent research has paid much attention to enhancing interpretability or explainability to allow machine to explain how it makes a decision for supporting humans, this research argues that there is urging the need for both human and AI should develop specific, corresponding ability to interact and collaborate with each other to form a human-AI team to accomplish superior results. This research introduces a conceptual framework called "Co-Learning," in which people can learn with/from and grow with AI partners over time. We characterize three key concepts of co-learning: "mutual understanding," "mutual benefits," and "mutual growth" for facilitating human-AI collaboration on complex problem solving. We will present proof-of-concepts to investigate whether and how our approach can help human-AI team to understand and benefit each other, and ultimately improve productivity and creativity on creative problem domains. The insights will contribute to the design of Human-AI collaboration.

Chih-Yuan Yang, Heeseung Yun, Srenavis Varadaraj et al.

We develop a system which generates summaries from seniors' indoor-activity videos captured by a social robot to help remote family members know their seniors' daily activities at home. Unlike the traditional video summarization datasets, indoor videos captured from a moving robot poses additional challenges, namely, (i) the video sequences are very long (ii) a significant number of video-frames contain no-subject or with subjects at ill-posed locations and scales (iii) most of the well-posed frames contain highly redundant information. To address this problem, we propose to \hl{exploit} pose estimation \hl{for detecting} people in frames\hl{. This guides the robot} to follow the user and capture effective videos. We use person identification to distinguish a target senior from other people. We \hl{also make use of} action recognition to analyze seniors' major activities at different moments, and develop a video summarization method to select diverse and representative keyframes as summaries.

Chieh-Teng Chang, Chi-Chia Huang, Chih-Yuan Yang et al.

Automatic abstractive text summarization is an important and challenging research topic of natural language processing. Among many widely used languages, the Chinese language has a special property that a Chinese character contains rich information comparable to a word. Existing Chinese text summarization methods, either adopt totally character-based or word-based representations, fail to fully exploit the information carried by both representations. To accurately capture the essence of articles, we propose a hybrid word-character approach (HWC) which preserves the advantages of both word-based and character-based representations. We evaluate the advantage of the proposed HWC approach by applying it to two existing methods, and discover that it generates state-of-the-art performance with a margin of 24 ROUGE points on a widely used dataset LCSTS. In addition, we find an issue contained in the LCSTS dataset and offer a script to remove overlapping pairs (a summary and a short text) to create a clean dataset for the community. The proposed HWC approach also generates the best performance on the new, clean LCSTS dataset.