Record soundsTrain personalized model Iteratively test Figure 1: Overview of the SPECTRA pipeline.In an interactive machine learning training workfow, users collect audio data samples (left), flter their data into a training dataset (center), and assess their model's performance in a live environment (right).The design includes key elements to support the needs of DHH users during this process, including spectrogram and waveform audio visualizations of audio, data annotating to save useful contextual information, and an interactive clustering visualization of their dataset.

With improvements in automated speech recognition and increased use of videoconferencing, real-time captioning has changed significantly. This shift toward broadly available but less accurate captioning invites exploration of the role hearing conversation partners play in shaping the accessibility of a conversation to d/Deaf and hard of hearing (DHH) captioning users. While recent work has explored DHH individuals’ videoconferencing experiences with captioning, we focus on established groups’ current practices and priorities for future tools to support more accessible online conversations. Our study consists of three codesign sessions, conducted with four groups (17 participants total, 10 DHH, 7 hearing). We found that established groups crafted social accessibility norms that met their relational contexts. We also identify promising directions for future captioning design, including the need to standardize speaker identification and customization, opportunities to provide behavioral feedback during a conversation, and ways that videoconferencing platforms could enable groups to set and share norms.

Prior work has explored the writing challenges experienced by people with dyslexia, and the potential for new spelling, grammar, and word retrieval technologies to address these challenges. However, the capabilities for natural language generation demonstrated by the latest class of large language models (LLMs) highlight an opportunity to explore new forms of human-AI writing support tools. In this paper, we introduce LaMPost, a prototype email-writing interface that explores the potential for LLMs to power writing support tools that address the varied needs of people with dyslexia. LaMPost draws from our understanding of these needs and introduces novel AI-powered features for email-writing, including: outlining main ideas, generating a subject line, suggesting changes, rewriting a selection. We evaluated LaMPost with 19 adults with dyslexia, identifying many promising routes for further exploration (including the popularity of the “rewrite” and “subject line” features), but also finding that the current generation of LLMs may not surpass the accuracy and quality thresholds required to meet the needs of writers with dyslexia. Surprisingly, we found that participants’ awareness of the AI had no effect on their perception of the system, nor on their feelings of autonomy, expression, and self-efficacy when writing emails. Our findings yield further insight into the benefits and drawbacks of using LLMs as writing support for adults with dyslexia and provide a foundation to build upon in future research.

Recent advances have enabled automatic sound recognition systems for deaf and hard of hearing (DHH) users on mobile devices. However, these tools use pre-trained, generic sound recognition models, which do not meet the diverse needs of DHH users. We introduce ProtoSound, an interactive system for customizing sound recognition models by recording a few examples, thereby enabling personalized and fine-grained categories. ProtoSound is motivated by prior work examining sound awareness needs of DHH people and by a survey we conducted with 472 DHH participants. To evaluate ProtoSound, we characterized performance on two real-world sound datasets, showing significant improvement over state-of-the-art (e.g., +9.7% accuracy on the first dataset). We then deployed ProtoSound's end-user training and real-time recognition through a mobile application and recruited 19 hearing participants who listened to the real-world sounds and rated the accuracy across 56 locations (e.g., homes, restaurants, parks). Results show that ProtoSound personalized the model on-device in real-time and accurately learned sounds across diverse acoustic contexts. We close by discussing open challenges in personalizable sound recognition, including the need for better recording interfaces and algorithmic improvements.

Real-time captioning is a critical accessibility tool for many d/Deaf and hard of hearing (DHH) people. While the vast majority of captioning work has focused on formal settings and technical innovations, in contrast, we investigate captioning for informal, interactive small-group conversations, which have a high degree of spontaneity and foster dynamic social interactions. This paper reports on semi-structured interviews and design probe activities we conducted with 15 DHH participants to understand their use of existing real-time captioning services and future design preferences for both in-person and remote small-group communication. We found that our participants' experiences of captioned small-group conversations are shaped by social, environmental, and technical considerations (e.g., interlocutors' pre-established relationships, the type of captioning displays available, and how far captions lag behind speech). When considering future captioning tools, participants were interested in greater feedback on non-speech elements of conversation (e.g., speaker identity, speech rate, volume) both for their personal use and to guide hearing interlocutors toward more accessible communication. We contribute a qualitative account of DHH people's real-time captioning experiences during small-group conversation and future design considerations to better support the groups being captioned, both in person and online.?

Head-mounted displays can provide private and glanceable speech and sound feedback to deaf and hard of hearing people, yet prior systems have largely focused on speech transcription. We introduce HoloSound, a HoloLens-based augmented reality (AR) prototype that uses deep learning to classify and visualize sound identity and location in addition to providing speech transcription. This poster paper presents a working proof-of-concept prototype, and discusses future opportunities for advancing AR-based sound awareness.

Smartwatches have the potential to provide glanceable, always-available sound feedback to people who are deaf or hard of hearing. In this paper, we present a performance evaluation of four low-resource deep learning sound classification models: MobileNet, Inception, ResNet-lite, and VGG-lite across four device architectures: watch-only, watch+phone, watch+phone+cloud, and watch+cloud. While direct comparison with prior work is challenging, our results show that the best model, VGG-lite, performed similar to the state of the art for non-portable devices with an average accuracy of 81.2% (SD=5.8%) across 20 sound classes and 97.6% (SD=1.7%) across the three highest-priority sounds. For device architectures, we found that the watch+phone architecture provided the best balance between CPU, memory, network usage, and classification latency. Based on these experimental results, we built and conducted a qualitative lab evaluation of a smartwatch-based sound awareness app, called SoundWatch (Figure 1), with eight DHH participants. Qualitative findings show support for our sound awareness app but also uncover issues with misclassifications, latency, and privacy concerns. We close by offering design considerations for future wearable sound awareness technology.

We present a qualitative study with 16 deaf and hard of hearing (DHH) participants examining reactions to smartwatch-based visual + haptic sound feedback designs. In Part 1, we conducted a Wizard-of-Oz (WoZ) evaluation of three smartwatch feedback techniques (visual alone, visual + simple vibration, and visual + tacton) and investigated vibrational patterns (tactons) to portray sound loudness, direction, and identity. In Part 2, we visited three public or semi-public locations where we demonstrated sound feedback on the smartwatch in situ to examine contextual influences and explore sound filtering options. Our findings characterize uses for vibration in multimodal sound awareness, both for push notification and for immediately actionable sound information displayed through vibrational patterns (tactons). In situ experiences caused participants to request sound filtering - particularly to limit haptic feedback - as a method for managing soundscape complexity. Additional concerns arose related to learnability, possibility of distraction, and system trust. Our findings have implications for future portable sound awareness systems.

We introduce HomeSound, an in-home sound awareness system for Deaf and hard of hearing (DHH) users. Similar to the Echo Show or Nest Hub, HomeSound consists of a microphone and display, and uses multiple devices installed in each home. We iteratively developed two prototypes, both of which sense and visualize sound information in real-time. Prototype 1 provided a floorplan view of sound occurrences with waveform histories depicting loudness and pitch. A three-week deployment in four DHH homes showed an increase in participants' home- and self-awareness but also uncovered challenges due to lack of line of sight and sound classification. For Prototype 2, we added automatic sound classification and smartwatch support for wearable alerts. A second field deployment in four homes showed further increases in awareness but misclassifications and constant watch vibrations were not well received. We discuss findings related to awareness, privacy, and display placement and implications for future home sound awareness technology.