arXiv subjects

Amy Pavel

Publications and source records attributed to Amy Pavel.

At least 19 recordsLinked to original sources

arXiv2026

Cyber-Physical Systems for Accessibility and Ability Augmentation: Bridging Diverse Communities

The powerful convergence of wearables, robotics, extended reality, and smart environments is expanding the design space for cyber-physical systems (CPS) that support and augment human abilities in daily life. By sensing real-world contexts, modeling user needs, and providing situated assistance, these systems can improve accessibility for people with disabilities while enhancing broader human abilities such as perception, memory, learning, and mobility. However, realizing this potential requires addressing key challenges in context sensing, user modeling, adaptive interaction, privacy, and evaluation to ensure that CPS are reliable and effective in real-world contexts. This workshop will bring together researchers and practitioners across HCI, AI, wearables, robotics, XR, smart environments, accessibility, and ability augmentation to examine shared strategies and challenges for designing accessibility- and ability-centered CPS. Through panel discussions, interactive demos, and mixed-group design activities, participants will identify recurring design principles, technical challenges, and future directions for CPS that support and augment human abilities in real-world settings. For details, please visit: https://cps4all.github.io.

cs.HC↗

arXiv2026

Dramarrator: Object-Based Audio Editing for Audio Drama Production from Books

Audio dramas weave dialogue, sound effects, and music into immersive stories. Creators often adapt books into audio dramas, but this process remains labor-intensive, requiring them to interpret source material, author scripts, generate audio assets, and assemble them on a timeline. Because story elements like characters and scenes manifest across many interdependent assets, a single change can ripple into manual updates across the entire project. We present Dramarrator, an audio drama authoring tool built around object-based audio editing, where these story elements are represented as editable objects. Dramarrator extracts these objects from a book, generates linked audio assets (speech, sound effects, and music), and composes a multi-track audio drama. Edits to any object (e.g., a character's voice) automatically propagate to all dependent assets. In a user study with professionals (N=8), Dramarrator significantly lowered task load when creating audio dramas. A listener study (N=300) shows that creator-refined output from Dramarrator approaches the quality of productions made with existing professional tools, and an exploratory study (N=3) suggests object-based editing lowers entry barriers and generalizes beyond audio dramas.

cs.HC↗

arXiv2026

TaskArtisan: Designing Composable Generative Widgets for LLM-Assisted Analysis

People increasingly use chatbots such as ChatGPT for everyday analysis tasks. While chatbots unify many analysis functions (e.g., scripts, visualizations, summaries), long conversations become hard to navigate, making it difficult to revisit prior steps or reuse successful workflows. LLMs now generate high-fidelity GUI code that enables people to create customized analysis tools beyond text. Yet, what new opportunities generative UIs bring to analysis work remain unclear. We interviewed six professionals about analysis with chatbots, analyzed publicly shared LLM-generated GUI tools, and conducted a comparison study (N=12) between a chatbot and TaskArtisan, a technology probe that enables people to create and assemble generative analysis UI widgets for sequential and fan-out composition. We find that GUI improved clarity and visual presentation but also introduced rigidity and additional prompting challenges. We summarize the trade-offs into a provisional design framework (malleability, specification, interoperability) to inform future generative UI in LLM-assisted analysis workflows.

cs.HC↗

arXiv2026

DigitalCoach: Communication and Grounding Gaps in Human and Agentic Computer Use Coaching

Agents are increasingly capable of automating software tasks, but can they teach humans how to use software themselves? We introduce DigitalCoach, a multimodal dataset of 72 human expert-novice computer use coaching sessions consisting of 22,752 dialogue turns grounded in 28.1 hours of screen and input event recordings across five software applications. We use DigitalCoach to evaluate whether state-of-the-art models can teach humans how to use computers. Automated evaluation shows that models differ from humans in how they coach: models provide more direct instructions, but fewer explanations, error diagnoses, and knowledge-check questions. When we fix the coaching method, models produce utterances similar to human references yet poorly grounded in visual context. Interactive evaluation confirms that model coaches cause learners to passively follow instructions without deeper engagement and fall short in visual grounding. DigitalCoach lays a foundation for collaborative and proactive computer use coaching agents. Data and code are available at https://project-digital-coach.vercel.app.

cs.CL↗

arXiv2026

A11y-CUA Dataset: Characterizing the Accessibility Gap in Computer Use Agents

Computer Use Agents (CUAs) operate interfaces by pointing, clicking, and typing -- mirroring interactions of sighted users (SUs) who can thus monitor CUAs and share control. CUAs do not reflect interactions by blind and low-vision users (BLVUs) who use assistive technology (AT). BLVUs thus cannot easily collaborate with CUAs. To characterize the accessibility gap of CUAs, we present A11y-CUA, a dataset of BLVUs and SUs performing 60 everyday tasks with 40.4 hours and 158,325 events. Our dataset analysis reveals that our collected interaction traces quantitatively confirm distinct interaction styles between SU and BLVU groups (mouse- vs. keyboard-dominant) and demonstrate interaction diversity within each group (sequential vs. shortcut navigation for BLVUs). We then compare collected traces to state-of-the-art CUAs under default and AT conditions (keyboard-only, magnifier). The default CUA executed 78.3% of tasks successfully. But with the AT conditions, CUA's performance dropped to 41.67% and 28.3% with keyboard-only and magnifier conditions respectively, and did not reflect nuances of real AT use. With our open A11y-CUA dataset, we aim to promote collaborative and accessible CUAs for everyone.

cs.HC↗

arXiv2025

Co-Designing Multimodal Systems for Accessible Asynchronous Dance Instruction

Videos make exercise instruction widely available, but they rely on visual demonstrations that blind and low vision (BLV) learners cannot see. While audio descriptions (AD) can make videos accessible, describing movements remains challenging as the AD must convey what to do (mechanics, location, orientation) and how to do it (speed, fluidity, timing). Prior work thus used multimodal instruction to support BLV learners with individual simple movements. However, it is unclear how these approaches scale to dance instruction with unique, complex movements and precise timing constraints. To inform accessible asynchronous dance instruction systems, we conducted three co-design workshops (N=28) with BLV dancers, instructors, and experts in sound, haptics, and AD. Participants designed 8 systems revealing common themes: staged learning to dissect routines, crafting vocabularies for movements, and selectively using modalities (narration for movement structure, sound for expression, and haptics for spatial cues). We conclude with design implications to make learning dance accessible.

cs.HC↗

arXiv2025

Morae: Proactively Pausing UI Agents for User Choices

User interface (UI) agents promise to make inaccessible or complex UIs easier to access for blind and low-vision (BLV) users. However, current UI agents typically perform tasks end-to-end without involving users in critical choices or making them aware of important contextual information, thus reducing user agency. For example, in our field study, a BLV participant asked to buy the cheapest available sparkling water, and the agent automatically chose one from several equally priced options, without mentioning alternative products with different flavors or better ratings. To address this problem, we introduce Morae, a UI agent that automatically identifies decision points during task execution and pauses so that users can make choices. Morae uses large multimodal models to interpret user queries alongside UI code and screenshots, and prompt users for clarification when there is a choice to be made. In a study over real-world web tasks with BLV participants, Morae helped users complete more tasks and select options that better matched their preferences, as compared to baseline agents, including OpenAI Operator. More broadly, this work exemplifies a mixed-initiative approach in which users benefit from the automation of UI agents while being able to express their preferences.

cs.HC↗

arXiv2025

CoSight: Exploring Viewer Contributions to Online Video Accessibility Through Descriptive Commenting

The rapid growth of online video content has outpaced efforts to make visual information accessible to blind and low vision (BLV) audiences. While professional Audio Description (AD) remains the gold standard, it is costly and difficult to scale across the vast volume of online media. In this work, we explore a complementary approach to broaden participation in video accessibility: engaging everyday video viewers at their watching and commenting time. We introduce CoSight, a Chrome extension that augments YouTube with lightweight, in-situ nudges to support descriptive commenting. Drawing from Fogg's Behavior Model, CoSight provides visual indicators of accessibility gaps, pop-up hints for what to describe, reminders to clarify vague comments, and related captions and comments as references. In an exploratory study with 48 sighted users, CoSight helped integrate accessibility contribution into natural viewing and commenting practices, resulting in 89% of comments including grounded visual descriptions. Follow-up interviews with four BLV viewers and four professional AD writers suggest that while such comments do not match the rigor of professional AD, they can offer complementary value by conveying visual context and emotional nuance for understanding the videos.

cs.HC↗

arXiv2025

VeasyGuide: Personalized Visual Guidance for Low-vision Learners on Instructor Actions in Presentation Videos

Instructors often rely on visual actions such as pointing, marking, and sketching to convey information in educational presentation videos. These subtle visual cues often lack verbal descriptions, forcing low-vision (LV) learners to search for visual indicators or rely solely on audio, which can lead to missed information and increased cognitive load. To address this challenge, we conducted a co-design study with three LV participants and developed VeasyGuide, a tool that uses motion detection to identify instructor actions and dynamically highlight and magnify them. VeasyGuide produces familiar visual highlights that convey spatial context and adapt to diverse learners and content through extensive personalization and real-time visual feedback. VeasyGuide reduces visual search effort by clarifying what to look for and where to look. In an evaluation with 8 LV participants, learners demonstrated a significant improvement in detecting instructor actions, with faster response times and significantly reduced cognitive load. A separate evaluation with 8 sighted participants showed that VeasyGuide also enhanced engagement and attentiveness, suggesting its potential as a universally beneficial tool.

cs.HC↗

arXiv2025

Surfacing Variations to Calibrate Perceived Reliability of MLLM-generated Image Descriptions

Multimodal large language models (MLLMs) provide new opportunities for blind and low vision (BLV) people to access visual information in their daily lives. However, these models often produce errors that are difficult to detect without sight, posing safety and social risks in scenarios from medication identification to outfit selection. While BLV MLLM users use creative workarounds such as cross-checking between tools and consulting sighted individuals, these approaches are often time-consuming and impractical. We explore how systematically surfacing variations across multiple MLLM responses can support BLV users to detect unreliable information without visually inspecting the image. We contribute a design space for eliciting and presenting variations in MLLM descriptions, a prototype system implementing three variation presentation styles, and findings from a user study with 15 BLV participants. Our results demonstrate that presenting variations significantly increases users' ability to identify unreliable claims (by 4.9x using our approach compared to single descriptions) and significantly decreases perceived reliability of MLLM responses. 14 of 15 participants preferred seeing variations of MLLM responses over a single description, and all expressed interest in using our system for tasks from understanding a tornado's path to posting an image on social media.

cs.HC↗

arXiv2025

TalkLess: Blending Extractive and Abstractive Speech Summarization for Editing Speech to Preserve Content and Style

Millions of people listen to podcasts, audio stories, and lectures, but editing speech remains tedious and time-consuming. Creators remove unnecessary words, cut tangential discussions, and even re-record speech to make recordings concise and engaging. Prior work automatically summarized speech by removing full sentences (extraction), but rigid extraction limits expressivity. AI tools can summarize then re-synthesize speech (abstraction), but abstraction strips the speaker's style. We present TalkLess, a system that flexibly combines extraction and abstraction to condense speech while preserving its content and style. To edit speech, TalkLess first generates possible transcript edits, selects edits to maximize compression, coverage, and audio quality, then uses a speech editing model to translate transcript edits into audio edits. TalkLess's interface provides creators control over automated edits by separating low-level wording edits (via the compression pane) from major content edits (via the outline pane). TalkLess achieves higher coverage and removes more speech errors than a state-of-the-art extractive approach. A comparison study (N=12) showed that TalkLess significantly decreased cognitive load and editing effort in speech editing. We further demonstrate TalkLess's potential in an exploratory study (N=3) where creators edited their own speech.

cs.HC↗

arXiv2025

Task Mode: Dynamic Filtering for Task-Specific Web Navigation using LLMs

Modern web interfaces are unnecessarily complex to use as they overwhelm users with excessive text and visuals unrelated to their current goals. This problem particularly impacts screen reader users (SRUs), who navigate content sequentially and may spend minutes traversing irrelevant elements before reaching desired information compared to vision users (VUs) who visually skim in seconds. We present Task Mode, a system that dynamically filters web content based on user-specified goals using large language models to identify and prioritize relevant elements while minimizing distractions. Our approach preserves page structure while offering multiple viewing modes tailored to different access needs. Our user study with 12 participants (6 VUs, 6 SRUs) demonstrates that our approach reduced task completion time for SRUs while maintaining performance for VUs, decreasing the completion time gap between groups from 2x to 1.2x. 11 of 12 participants wanted to use Task Mode in the future, reporting that Task Mode supported completing tasks with less effort and fewer distractions. This work demonstrates how designing new interactions simultaneously for visual and non-visual access can reduce rather than reinforce accessibility disparities in future technology created by human-computer interaction researchers and practitioners.

cs.HC↗

arXiv2025

Vid2Coach: Transforming How-To Videos into Task Assistants

People use videos to learn new recipes, exercises, and crafts. Such videos remain difficult for blind and low vision (BLV) people to follow as they rely on visual comparison. Our observations of visual rehabilitation therapists (VRTs) guiding BLV people to follow how-to videos revealed that VRTs provide both proactive and responsive support including detailed descriptions, non-visual workarounds, and progress feedback. We propose Vid2Coach, a system that transforms how-to videos into wearable camera-based assistants that provide accessible instructions and mixed-initiative feedback. From the video, Vid2Coach generates accessible instructions by augmenting narrated instructions with demonstration details and completion criteria for each step. It then uses retrieval-augmented-generation to extract relevant non-visual workarounds from BLV-specific resources. Vid2Coach then monitors user progress with a camera embedded in commercial smart glasses to provide context-aware instructions, proactive feedback, and answers to user questions. BLV participants (N=8) using Vid2Coach completed cooking tasks with 58.5\% fewer errors than when using their typical workflow and wanted to use Vid2Coach in their daily lives. Vid2Coach demonstrates an opportunity for AI visual assistance that strengthens rather than replaces non-visual expertise.

cs.HC↗

arXiv2025

VideoDiff: Human-AI Video Co-Creation with Alternatives

To make an engaging video, people sequence interesting moments and add visuals such as B-rolls or text. While video editing requires time and effort, AI has recently shown strong potential to make editing easier through suggestions and automation. A key strength of generative models is their ability to quickly generate multiple variations, but when provided with many alternatives, creators struggle to compare them to find the best fit. We propose VideoDiff, an AI video editing tool designed for editing with alternatives. With VideoDiff, creators can generate and review multiple AI recommendations for each editing process: creating a rough cut, inserting B-rolls, and adding text effects. VideoDiff simplifies comparisons by aligning videos and highlighting differences through timelines, transcripts, and video previews. Creators have the flexibility to regenerate and refine AI suggestions as they compare alternatives. Our study participants (N=12) could easily compare and customize alternatives, creating more satisfying results.

cs.HC↗

arXiv2025

Lotus: Creating Short Videos From Long Videos With Abstractive and Extractive Summarization

Short-form videos are popular on platforms like TikTok and Instagram as they quickly capture viewers' attention. Many creators repurpose their long-form videos to produce short-form videos, but creators report that planning, extracting, and arranging clips from long-form videos is challenging. Currently, creators make extractive short-form videos composed of existing long-form video clips or abstractive short-form videos by adding newly recorded narration to visuals. While extractive videos maintain the original connection between audio and visuals, abstractive videos offer flexibility in selecting content to be included in a shorter time. We present Lotus, a system that combines both approaches to balance preserving the original content with flexibility over the content. Lotus first creates an abstractive short-form video by generating both a short-form script and its corresponding speech, then matching long-form video clips to the generated narration. Creators can then add extractive clips with an automated method or Lotus's editing interface. Lotus's interface can be used to further refine the short-form video. We compare short-form videos generated by Lotus with those using an extractive baseline method. In our user study, we compare creating short-form videos using Lotus to participants' existing practice.

cs.HC↗

arXiv2025

HistoryPalette: Supporting Exploration and Reuse of Past Alternatives in Image Generation and Editing

Creative tasks require creators to iteratively produce, select, and discard potentially useful ideas. Now, creativity tools include generative AI features (e.g., Photoshop Generative Fill) that increase the number of alternatives creators consider through rapid experiments with prompts and random generations. Creators use tedious manual systems for organizing their prior ideas by saving file versions or hiding layers, but they lack the support they want for reusing prior alternatives in personal work or in communication with others. We present HistoryPalette, a system that supports exploration and reuse of prior designs in generative image creation and editing. Using HistoryPalette, creators and their collaborators explore a "palette" of prior design alternatives organized by spatial position, topic category, and creation time. HistoryPalette enables creators to quickly preview and reuse their prior work. In creative professional and client collaborator user studies, participants generated and edited images by exploring and reusing past design alternatives with HistoryPalette.

cs.HC↗

arXiv2024

DreamStruct: Understanding Slides and User Interfaces via Synthetic Data Generation

Enabling machines to understand structured visuals like slides and user interfaces is essential for making them accessible to people with disabilities. However, achieving such understanding computationally has required manual data collection and annotation, which is time-consuming and labor-intensive. To overcome this challenge, we present a method to generate synthetic, structured visuals with target labels using code generation. Our method allows people to create datasets with built-in labels and train models with a small number of human-annotated examples. We demonstrate performance improvements in three tasks for understanding slides and UIs: recognizing visual elements, describing visual content, and classifying visual content types.

cs.CV↗

arXiv2024

Context-Aware Image Descriptions for Web Accessibility

Blind and low vision (BLV) internet users access images on the web via text descriptions. New vision-to-language models such as GPT-V, Gemini, and LLaVa can now provide detailed image descriptions on-demand. While prior research and guidelines state that BLV audiences' information preferences depend on the context of the image, existing tools for accessing vision-to-language models provide only context-free image descriptions by generating descriptions for the image alone without considering the surrounding webpage context. To explore how to integrate image context into image descriptions, we designed a Chrome Extension that automatically extracts webpage context to inform GPT-4V-generated image descriptions. We gained feedback from 12 BLV participants in a user study comparing typical context-free image descriptions to context-aware image descriptions. We then further evaluated our context-informed image descriptions with a technical evaluation. Our user evaluation demonstrated that BLV participants frequently prefer context-aware descriptions to context-free descriptions. BLV participants also rated context-aware descriptions significantly higher in quality, imaginability, relevance, and plausibility. All participants shared that they wanted to use context-aware descriptions in the future and highlighted the potential for use in online shopping, social media, news, and personal interest blogs.

cs.HC↗