Search

United Nations practice shows that inclusivity is vital for mediation to be successful in helping end violent conflict and establish lasting peace. However, current methods for understanding the views and needs of populations during dynamic situations create tension between inclusivity and efficiency. This work introduces a novel paradigm to mitigate such tension. In partnership with collaborators at the United Nations we develop a realtime large-scale synchronous dialogue process (RLSDP) to understand stakeholder populations on an hour timescale. We demonstrate a machine learning model which enables each dialogue cycle to take place on a minute-timescale. We manage a key risk related to machine learning result trustworthiness by computing result confidence from a fast and reliable estimation of posterior variance. Lastly, we highlight a constellation of risks stemming from this new paradigm and suggest policies to mitigate them., Comment: AI for Social Good workshop at NeurIPS (2019), Vancouver, Canada

Graphs have emerged as a natural choice to represent and analyze the intricate patterns and rich information of the Web, enabling applications such as online page classification and social recommendation. The prevailing "pre-train, fine-tune" paradigm has been widely adopted in graph machine learning tasks, particularly in scenarios with limited labeled nodes. However, this approach often exhibits a misalignment between the training objectives of pretext tasks and those of downstream tasks. This gap can result in the "negative transfer" problem, wherein the knowledge gained from pre-training adversely affects performance in the downstream tasks. The surge in prompt-based learning within Natural Language Processing (NLP) suggests the potential of adapting a "pre-train, prompt" paradigm to graphs as an alternative. However, existing graph prompting techniques are tailored to homogeneous graphs, neglecting the inherent heterogeneity of Web graphs. To bridge this gap, we propose HetGPT, a general post-training prompting framework to improve the predictive performance of pre-trained heterogeneous graph neural networks (HGNNs). The key is the design of a novel prompting function that integrates a virtual class prompt and a heterogeneous feature prompt, with the aim to reformulate downstream tasks to mirror pretext tasks. Moreover, HetGPT introduces a multi-view neighborhood aggregation mechanism, capturing the complex neighborhood structure in heterogeneous graphs. Extensive experiments on three benchmark datasets demonstrate HetGPT's capability to enhance the performance of state-of-the-art HGNNs on semi-supervised node classification., Comment: Accepted to WWW 2024 as research paper

Bipolar Disorder (BD) is a psychiatric condition diagnosed by repetitive cycles of hypomania and depression. Since diagnosing BD relies on subjective behavioral assessments over a long period, a solid diagnosis based on objective criteria is not straightforward. The current study responded to the described obstacle by proposing a hybrid GAN-CNN model to diagnose BD from 3-D structural MRI Images (sMRI). The novelty of this study stems from diagnosing BD from sMRI samples rather than conventional datasets such as functional MRI (fMRI), electroencephalography (EEG), and behavioral symptoms while removing the data insufficiency usually encountered when dealing with sMRI samples. The impact of various augmentation ratios is also tested using 5-fold cross-validation. Based on the results, this study obtains an accuracy rate of 75.8%, a sensitivity of 60.3%, and a specificity of 82.5%, which are 3-5% higher than prior work while utilizing less than 6% sample counts. Next, it is demonstrated that a 2- D layer-based GAN generator can effectively reproduce complex 3D brain samples, a more straightforward technique than manual image processing. Lastly, the optimum augmentation threshold for the current study using 172 sMRI samples is 50%, showing the applicability of the described method for larger sMRI datasets. In conclusion, it is established that data augmentation using GAN improves the accuracy of the CNN classifier using sMRI samples, thus developing more reliable decision support systems to assist practitioners in identifying BD patients more reliably and in a shorter period

Serving deep learning (DL) models on relational data has become a critical requirement across diverse commercial and scientific domains, sparking growing interest recently. In this visionary paper, we embark on a comprehensive exploration of representative architectures to address the requirement. We highlight three pivotal paradigms: The state-of-the-art DL-centric architecture offloads DL computations to dedicated DL frameworks. The potential UDF-centric architecture encapsulates one or more tensor computations into User Defined Functions (UDFs) within the relational database management system (RDBMS). The potential relation-centric architecture aims to represent a large-scale tensor computation through relational operators. While each of these architectures demonstrates promise in specific use scenarios, we identify urgent requirements for seamless integration of these architectures and the middle ground in-between these architectures. We delve into the gaps that impede the integration and explore innovative strategies to close them. We present a pathway to establish a novel RDBMS for enabling a broad class of data-intensive DL inference applications., Comment: * Authors are ordered alphabetically; Jia Zou is the corresponding author

Getting precise aspects of road through segmentation from remote sensing imagery is useful for many real-world applications such as autonomous vehicles, urban development and planning, and achieving sustainable development goals. Roads are only a small part of the image, and their appearance, type, width, elevation, directions, etc. exhibit large variations across geographical areas. Furthermore, due to differences in urbanization styles, planning, and the natural environments; regions along the roads vary significantly. Due to these variations among the train and test domains, the road segmentation algorithms fail to generalize to new geographical locations. Unlike the generic domain alignment scenarios, road segmentation has no scene structure, and generic domain adaptation methods are unable to enforce topological properties like continuity, connectivity, smoothness, etc., thus resulting in degraded domain alignment. In this work, we propose a topology-aware unsupervised domain adaptation approach for road segmentation in remote sensing imagery. Specifically, we predict road skeleton, an auxiliary task to impose the topological constraints. To enforce consistent predictions of road and skeleton, especially in the unlabeled target domain, the conformity loss is defined across the skeleton prediction head and the road-segmentation head. Furthermore, for self-training, we filter out the noisy pseudo-labels by using a connectivity-based pseudo-labels refinement strategy, on both road and skeleton segmentation heads, thus avoiding holes and discontinuities. Extensive experiments on the benchmark datasets show the effectiveness of the proposed approach compared to existing state-of-the-art methods. Specifically, for SpaceNet to DeepGlobe adaptation, the proposed approach outperforms the competing methods by a minimum margin of 6.6%, 6.7%, and 9.8% in IoU, F1-score, and APLS, respectively.

Noise robustness is a key aspect of successful speech applications. Speech enhancement (SE) has been investigated to improve automatic speech recognition accuracy; however, its effectiveness for keyword spotting (KWS) is still under-investigated. In this paper, we conduct a comprehensive study on single-channel speech enhancement for keyword spotting on the Google Speech Command (GSC) dataset. To investigate robustness to noise, the GSC dataset is augmented with noise signals from the WSJ0 Hipster Ambient Mixtures (WHAM!) noise dataset. Our investigation includes not only applying SE before KWS but also performing joint training of the SE frontend and KWS backend models. Moreover, we explore audio injection, a common approach to reduce distortions by using a weighted average of the enhanced and original signals. Audio injection is then further optimized by using another model that predicts the weight for each utterance. Our investigation reveals that SE can improve KWS accuracy on noisy speech when the backend model is trained on clean speech; however, despite our extensive exploration, it is difficult to improve the KWS accuracy with SE when the backend is trained on noisy speech.

Voice triggering (VT) enables users to activate their devices by just speaking a trigger phrase. A front-end system is typically used to perform speech enhancement and/or separation, and produces multiple enhanced and/or separated signals. Since conventional VT systems take only single-channel audio as input, channel selection is performed. A drawback of this approach is that unselected channels are discarded, even if the discarded channels could contain useful information for VT. In this work, we propose multichannel acoustic models for VT, where the multichannel output from the frond-end is fed directly into a VT model. We adopt a transform-average-concatenate (TAC) block and modify the TAC block by incorporating the channel from the conventional channel selection so that the model can attend to a target speaker when multiple speakers are present. The proposed approach achieves up to 30% reduction in the false rejection rate compared to the baseline channel selection approach., Comment: Accepted at HSCMA 2024

Cough acoustics contain multitudes of vital information about pathomorphological alterations in the respiratory system. Reliable and accurate detection of cough events by investigating the underlying cough latent features and disease diagnosis can play an indispensable role in revitalizing the healthcare practices. The recent application of Artificial Intelligence (AI) and advances of ubiquitous computing for respiratory disease prediction has created an auspicious trend and myriad of future possibilities in the medical domain. In particular, there is an expeditiously emerging trend of Machine learning (ML) and Deep Learning (DL)-based diagnostic algorithms exploiting cough signatures. The enormous body of literature on cough-based AI algorithms demonstrate that these models can play a significant role for detecting the onset of a specific respiratory disease. However, it is pertinent to collect the information from all relevant studies in an exhaustive manner for the medical experts and AI scientists to analyze the decisive role of AI/ML. This survey offers a comprehensive overview of the cough data-driven ML/DL detection and preliminary diagnosis frameworks, along with a detailed list of significant features. We investigate the mechanism that causes cough and the latent cough features of the respiratory modalities. We also analyze the customized cough monitoring application, and their AI-powered recognition algorithms. Challenges and prospective future research directions to develop practical, robust, and ubiquitous solutions are also discussed in detail., Comment: 30 pages, 12 figures, 9 tables

Despite constituting a noteworthy $\sim 27\%$ share of the total energy budget of our Universe, dark matter (DM) has thus far eluded direct observations. Owing to its pervasive nature, there is a sincere expectation that astrophysical black holes (BHs) encompassed by DM should leave distinctive imprints on the gravitational waves arising from BH mergers. Theoretical models of DM present a diverse landscape of possibilities, with perfect fluid dark matter (PFDM) emerging as a recent and notably intriguing candidate model. In this work, utilizing the established quantum optical approach, we investigate the possibility of catching DM signatures via acceleration radiation emitted by a freely-falling detector (e.g. an atom) within a PFDM-surrounded Schwarzschild BH. The setup involves a Casimir-type apparatus where the detector interacts with the field, and this situation induces excitations in the detector in a manner consistent with Unruh effect. We observe that our DM candidate, while making classical contributions to spacetime geometry, has the potential to leave quantum imprints in the radiation flux. Notably, it is observed that, in comparison to a pure Schwarzschild BH, PFDM can markedly reduce particle emission as long as its density remains below a critical threshold, and vice versa. Given the lessons we have learnt from realizing cosmological phenomena in simulated laboratory conditions, there is a remote possibility that such study may perhaps provide insights (to whatever degree !) into the future table-top experiments in analogue gravity paradigm., Comment: 8 pages, 6 figures