Advancements in audio neural networks have established state-of-the-art results on downstream audio tasks. However, the black-box structure of these models makes it difficult to interpret the information encoded in their internal audio representations. In this work, we explore the semantic interpretability of audio embeddings extracted from these neural networks by leveraging CLAP, a contrastive learning model that brings audio and text into a shared embedding space. We implement a post-hoc method to transform CLAP embeddings into concept-based, sparse representations with semantic interpretability. Qualitative and quantitative evaluations show that the concept-based representations outperform or match the performance of original audio embeddings on downstream tasks while providing interpretability. Additionally, we demonstrate that fine-tuning the concept-based representations can further improve their performance on downstream tasks. Lastly, we publish three audio-specific vocabularies for concept-based interpretability of audio embeddings.

Interactive voice assistants have been widely used as input interfaces in various scenarios, e.g. on smart home devices, wearables and on AR devices. Detecting the end of a speech query, i.e. speech end-pointing, is an important task for voice assistants to interact with users. Traditionally, speech end-pointing is based on pure classification methods along with arbitrary binary targets. In this paper, we propose a novel regression-based speech end-pointing model, which enables an end-pointer to adjust its detection behavior based on the context of user queries. Specifically, we present a pause modeling method and show its effectiveness for dynamic end-pointing. Based on our experiments with vendor-collected smartphone and wearables speech queries, our strategy shows a better trade-off between end-pointing latency and accuracy, compared to the traditional classification-based method. We further discuss the benefits of this model and generalization of the framework in the paper.

Digital platforms, including online forums and helplines, have emerged as avenues of support for caregivers suffering from postpartum mental health distress. Understanding support seekers' experiences as shared on these platforms could provide crucial insight into caregivers' needs during this vulnerable time. In the current work, we provide a descriptive analysis of the concerns, psychological states, and motivations shared by healthy and distressed postpartum support seekers on two digital platforms, a one-on-one digital helpline and a publicly available online forum. Using a combination of human annotations, dictionary models and unsupervised techniques, we find stark differences between the experiences of distressed and healthy mothers. Distressed mothers described interpersonal problems and a lack of support, with 8.60% - 14.56% reporting severe symptoms including suicidal ideation. In contrast, the majority of healthy mothers described childcare issues, such as questions about breastfeeding or sleeping, and reported no severe mental health concerns. Across the two digital platforms, we found that distressed mothers shared similar content. However, the patterns of speech and affect shared by distressed mothers differed between the helpline vs. the online forum, suggesting the design of these platforms may shape meaningful measures of their support-seeking experiences. Our results provide new insight into the experiences of caregivers suffering from postpartum mental health distress. We conclude by discussing methodological considerations for understanding content shared by support seekers and design considerations for the next generation of support tools for postpartum parents.

Most existing cry detection models have been tested with data collected in controlled settings. Thus, the extent to which they generalize to noisy and lived environments is unclear. In this paper, we evaluate several established machine learning approaches including a model leveraging both deep spectrum and acoustic features. This model was able to recognize crying events with F1 score 0.613 (Precision: 0.672, Recall: 0.552), showing improved external validity over existing methods at cry detection in everyday real-world settings. As part of our evaluation, we collect and annotate a novel dataset of infant crying compiled from over 780 hours of labeled real-world audio data, captured via recorders worn by infants in their homes, which we make publicly available. Our findings confirm that a cry detection model trained on in-lab data underperforms when presented with real-world data (in-lab test F1: 0.656, real-world test F1: 0.236), highlighting the value of our new dataset and model.

Audio has been increasingly adopted as a sensing modality in a variety of human-centered mobile applications and in smart assistants in the home. Although acoustic features can capture complex semantic information about human activities and context, continuous audio recording often poses significant privacy concerns. An intuitive way to reduce privacy concerns is to degrade audio quality such that speech and other relevant acoustic markers become unintelligible, but this often comes at the cost of activity recognition performance. In this paper, we employ a mixed-methods approach to characterize this balance. We first conduct an online survey with 266 participants to capture their perception of privacy qualitatively and quantitatively with degraded audio. Given our findings that privacy concerns can be significantly reduced at high levels of audio degradation, we then investigate how intentional degradation of audio frames can affect the recognition results of the target classes while maintaining effective privacy mitigation. Our results indicate that degradation of audio frames can leave minimal effects for audio recognition using frame-level features. Furthermore, degradation of audio frames can hurt the performance to some extend for audio recognition using segment-level features, though the usage of such features may still yield superior recognition performance. Given the different requirements on privacy mitigation and recognition performance for different sensing purposes, such trade-offs need to be balanced in actual implementations.

Over the last decade, advances in mobile technologies have enabled the development of intelligent systems that attempt to recognize and model a variety of health-related human behaviors. While automated dietary monitoring based on passive sensors has been an area of increasing research activity for many years, much less attention has been given to tracking fluid intake. In this work, we apply an adaptive segmentation technique on a continuous stream of inertial data captured with a practical, off-the-shelf wrist-mounted device to detect fluid intake gestures passively. We evaluated our approach in a study with 30 participants where 561 drinking instances were recorded. Using a leave-one-participant-out (LOPO), we were able to detect drinking episodes with 90.3% precision and 91.0% recall, demonstrating the generalizability of our approach. In addition to our proposed method, we also contribute an anonymized and labeled dataset of drinking and non-drinking gestures to encourage further work in the field.

The field of mobile health (mHealth) has the potential to yield new insights into health and behavior through the analysis of continuously recorded data from wearable health and activity sensors. In this paper, we present a hierarchical span-based conditional random field model for the key problem of jointly detecting discrete events in such sensor data streams and segmenting these events into high-level activity sessions. Our model includes higher-order cardinality factors and inter-event duration factors to capture domain-specific structure in the label space. We show that our model supports exact MAP inference in quadratic time via dynamic programming, which we leverage to perform learning in the structured support vector machine framework. We apply the model to the problems of smoking and eating detection using four real data sets. Our results show statistically significant improvements in segmentation performance relative to a hierarchical pairwise CRF.

Although food journaling is understood to be both important and difficult, little work has empirically documented the specific challenges people experience with food journals. We identify key challenges in a qualitative study combining a survey of 141 current and lapsed food journalers with analysis of 5,526 posts in community forums for three mobile food journals. Analyzing themes in this data, we find and discuss barriers to reliable food entry, negative nudges caused by current techniques, and challenges with social features. Our results motivate research exploring a wider range of approaches to food journal design and technology.

Recognizing when eating activities take place is one of the key challenges in automated food intake monitoring. Despite progress over the years, most proposed approaches have been largely impractical for everyday usage, requiring multiple on-body sensors or specialized devices such as neck collars for swallow detection. In this paper, we describe the implementation and evaluation of an approach for inferring eating moments based on 3-axis accelerometry collected with a popular off-the-shelf smartwatch. Trained with data collected in a semi-controlled laboratory setting with 20 subjects, our system recognized eating moments in two free-living condition studies (7 participants, 1 day; 1 participant, 31 days), with F-scores of 76.1% (66.7% Precision, 88.8% Recall), and 71.3% (65.2% Precision, 78.6% Recall). This work represents a contribution towards the implementation of a practical, automated system for everyday food intake monitoring, with applicability in areas ranging from health research and food journaling.

We explore using the Outer Ear Interface (OEI) to recognize eating activities. OEI contains a 3D gyroscope and a set of proximity sensors encapsulated in an off-the-shelf earpiece to monitor jaw movement by measuring ear canal deformation. In a laboratory setting with 20 participants, OEI could distinguish eating from other activities, such as walking, talking, and silently reading, with over 90% accuracy (user independent). In a second study, six subjects wore the system for 6 hours each while performing their normal daily activities. OEI correctly classified five minute segments of time as eating or non-eating with 93% accuracy (user dependent).

Dietary self-monitoring has been shown to be an effective method for weight-loss, but it remains an onerous task despite recent advances in food journaling systems. Semi-automated food journaling can reduce the effort of logging, but often requires that eating activities be detected automatically. In this work we describe results from a feasibility study conducted in-the-wild where eating activities were inferred from ambient sounds captured with a wrist-mounted device; twenty participants wore the device during one day for an average of 5 hours while performing normal everyday activities. Our system was able to identify meal eating with an F-score of 79.8% in a person-dependent evaluation, and with 86.6% accuracy in a person-independent evaluation. Our approach is intended to be practical, leveraging off-the-shelf devices with audio sensing capabilities in contrast to systems for automated dietary assessment based on specialized sensors.

The pervasiveness of mobile cameras has resulted in a dramatic increase in food photos, which are pictures reflecting what people eat. In this paper, we study how taking pictures of what we eat in restaurants can be used for the purpose of automating food journaling. We propose to leverage the context of where the picture was taken, with additional information about the restaurant, available online, coupled with state-of-the-art computer vision techniques to recognize the food being consumed. To this end, we demonstrate image-based recognition of foods eaten in restaurants by training a classifier with images from restaurant's online menu databases. We evaluate the performance of our system in unconstrained, real-world settings with food images taken in 10 restaurants across 5 different types of food (American, Indian, Italian, Mexican and Thai).

First-person point-of-view (FPPOV) images taken by wearable cameras can be used to better understand people's eating habits. Human computation is a way to provide effective analysis of FPPOV images in cases where algorithmic approaches currently fail. However, privacy is a serious concern. We provide a framework, the privacy-saliency matrix, for understanding the balance between the eating information in an image and its potential privacy concerns. Using data gathered by 5 participants wearing a lanyard-mounted smartphone, we show how the framework can be used to quantitatively assess the effectiveness of four automated techniques (face detection, image cropping, location filtering and motion filtering) at reducing the privacy-infringing content of images while still maintaining evidence of eating behaviors throughout the day.

Activity recognition in the home has been long recognized as the foundation for many desirable applications in fields such as home automation, sustainability, and healthcare. However, building a practical home activity monitoring system remains a challenge. Striking a balance between cost, privacy, ease of installation and scalability continues to be an elusive goal. In this paper, we explore infrastructure-mediated sensing combined with a vector space model learning approach as the basis of an activity recognition system for the home. We examine the performance of our single-sensor water-based system in recognizing eleven high-level activities in the kitchen and bathroom, such as cooking and shaving. Results from two studies show that our system can estimate activities with overall accuracy of 82.69% for one individual and 70.11% for a group of 23 participants. As far as we know, our work is the first to employ infrastructure-mediated sensing for inferring high-level human activities in a home setting.