With the rapid advancement in both medicine and technology, the lives of individuals with hearing impairments have witnessed significant improvements.

Nevertheless, in the course of our investigation, it has come to light that in certain settings, such as hair salons, individuals with hearing impairments are compelled to remove their hearing aids. Consequently, this poses a considerable challenge in effective communication with hairstylists.

With the rapid advancement of Artificial Intelligence(AI) technology, particularly in the realm of voice recognition, there has been notable progress in supporting individuals with hearing impairments.

This paper introduces the design of a system aimed at providing an improved communication experience for individuals with hearing impairments in hair salon environments.

An experiment were conducted to evaluate the performance of the system. The experiment included two tasks. We recruited 8 participants included 2 hairstylists and 6 hearing impaired. We conducted four questionnaires and a semi-structured interview.

In the future, the system can be integrated with other existing IoT technologies for use in real hair salons and can be extended to broader users beyond the hearing impaired, including foreigners.

It is difficult to be aware of the auditory signals of everyday life when one has a hearing impairment. In this study, we investigated which everyday acoustic signals are difficult for young deaf and hard of hearing (D/HoH) people to listen to. The participants were fifteen university students between the ages of 19 and 22. The auditory background varied (hearing aids, cochlear implants, bilateral hearing loss). Listening tests were conducted at different signal-to-noise ratios (SNR). The materials were clean sounds without noise and recorded sounds with noise (recorded in Taiwan and Japan). The materials were presented from loudspeakers in the classroom. The loudness of sounds was approximately 75 dBA at 1 m for the nearest person. Participants were asked to write down the name of the sound. The degrees of familiarity, confidence, and awareness were also assessed using a 5-point Mean Opinion Score (MOS). The results showed that the correct response rate tended to be high for sounds with good SNR. Significant correlations with the correct response rate were found for all three scores. From this result, we found that even if they thought the sound was easy to hear or they had confidence, they still listened to it as the wrong sound. For unfamiliar sounds, the percentage of correct responses is low. In this way, it is difficult for D/HoH people to passively recognize environmental sounds, and it is important to understand the environmental sounds around them through learning.

Assessing the comprehension of Deaf and Hard of Hearing (D/HoH) viewers in the context of television content has traditionally relied on methods such as focus groups, open-ended questionnaires, and multiple choice questions. This paper proposes a novel methodology rooted in the social pragmatic model of conversation to measure the comprehension of D/HoH viewers who watch fast-paced live sports. Sixteen Deaf and 11 Hard of Hearing individuals, all sports fans, engaged in one-on-one informal conversations with researchers after viewing a sports game with two different styles of captioning. Comprehension scores were determined based on correct responses to predetermined topics related to gameplay and commentary. Results indicate comprehension scores were above 60% with the majority of participants exceeding 80%. This conversation methodology was effective in capturing participants' understanding beyond traditional comprehension measures, encouraging spontaneous engagement and insights. Researchers recognized the natural and free-flowing nature of the conversations. This study introduces a promising method for evaluating content comprehension in D/HoH viewers, emphasizing the need for researchers to immerse themselves in content and conversations. Future research should explore variations in this methodology and topic selection while addressing training needs for researchers.

Canadian regulations require broadcasters to report the quality of Closed Captioning (CC) by using the Number-Edition-Recognition (NER) model, which requires certified human evaluators to assess CC in English programming. However, current methods to perform these evaluations are inefficient and lack a dedicated tool. This paper explores the development of NERWEB, a dedicated software tool designed to improve the efficiency and user experience of CC quality evaluation under the Canadian NER model. Employing a User-Centered Design approach, we conducted interviews and user studies with five certified evaluators. The resulting prototype offers a more integrated, efficient, and user-friendly interface compared to traditional methods. We present a set of design recommendations for developing a software interface dedicated to performing the NER quality evaluation.

Closed sign language interpreting (SLI) makes media accessible to deaf and hard of hearing viewers who use sign languages as their primary mode of communication. Analogous to subtitles for the deaf and hard of hearing, this feature allows to toggle sign language interpretation on and off, and customize its appearance in conjunction with videos. This paper provides information on how to design closed interpreting in a media player through a pair of mixed-method studies. The first study assesses the usability of technical SLI features, while the second one assesses how users interact with the content and focus their attention via eye gaze. Results indicate above-average usability for the technical features. Additionally, preliminary results on eye gaze suggest that optimal configuration of the SLI depends on the type of content viewed, and that user preferences vary. Overall, customizability of features and placement will be important in closed interpreting implementations.