Beyond the "Industry Standard": Focusing Gender-Affirming Voice Training Technologies on Individualized Goal Exploration

Research Background and Issues

• Identified Problems and Challenges:
  The authors point out that current gender-affirming voice training technologies often adhere to binary gender standards, neglecting the complex needs of transgender individuals in identity expression and personalized goal exploration. This design approach results in insufficient personalization and hinders transgender individuals from achieving more natural voice expressions aligned with their self-identity.
• Significance:
  Gender-affirming voice training is crucial for transgender individuals as it reduces gender dysphoria, enhances social integration, and ensures consistency in identity expression. This consistency not only impacts psychological well-being but also influences safety and confidence in social interactions. However, current technologies fail to adequately address these diverse needs.
• Research Motivation and Related Work:
  The authors emphasize their departure from existing work by seeking to comprehensively understand the dynamic process of personalized goal-setting in voice exploration, rather than simplifying voice goals to fixed standards measured by binary gender pitch ranges. They draw on prior research in human-computer interaction (HCI), voice training technologies, and gender expression technologies to critique current designs and advocate for the development of more community-inclusive technologies.

Solution

• Proposed Methods or Solutions:
  Through interviews with six voice training experts and ten transgender individuals with voice training experience, the authors explore how personalized voice goals are defined, tracked, and adjusted. Based on participant feedback, they propose a new framework for designing gender-affirming voice technologies. The study emphasizes separating goal-setting from practice and supporting personalized goal exploration through tools.
• Innovations:
  1. Highlighting the duality of voice goals: distinguishing between "descriptive goals" (expressing social identity and cultural needs) and "technical goals" (specific training objectives based on acoustic parameters).
  2. Proposing a dynamic process for goal exploration and reevaluation: bridging the gap between trainees' aspirations and achievable goals through role-playing, voice samples, and technical feedback.
  3. Exploring the potential of AI-based real-time voice modifiers and virtual reality (VR) technologies in personalized voice goal exploration.
• Implementation Steps and Technologies:
  • Goal Exploration: Providing a voice sample database for users to filter and match voice goals that align with their identity expression, using detailed parameters such as culture, age, and gender expression.
  • Goal Refinement: Generating new personalized voice samples using enhanced voice conversion (e.g., parameter-based acoustic fine-tuning) and retrieval-based voice conversion (RVC) technologies, guiding users to create unique voices through comparison and modification.
  • Dynamic Feedback Mechanism for Goal Achievement: Regularly reevaluating users' voice progress and offering nonlinear training pathways guided by goals, such as visual models illustrating the trajectory from initial voice to target voice.
  • Voice Practice Expansion: Utilizing social VR to provide simulated scenarios, helping users expand voice application contexts and reduce external environmental pressures.

Research Findings

• Specific Outcomes:
  1. Establishing the dynamic nature of voice goals: goals evolve with the growth of voice knowledge, skills, and deepening individual identity recognition.
  2. Clarifying the transformation between descriptive goals and technical goals, achievable through a combination of voice samples, role-playing, and parameter exploration.
  3. Developing a comprehensive technical design framework to support diverse voice goal exploration, including an integrated voice database, improved voice modifiers, and VR-based safe practice environments.
• Advantages Compared to Existing Solutions:
  • Centered on individual identity needs rather than imposing fixed binary gender pitch references.
  • Rich tools and methods enable greater flexibility in goal exploration, such as example-driven and dynamic adjustments.
  • Emphasizing the gradual establishment of "voice ownership" for the first time in voice technology applications, enhancing users' sense of self-fulfillment.
• Experimental and Evaluation Results:
  Through participatory interviews, consistent coding systems, and inductive analysis, the study found that successful voice training involves the combination of multidimensional goal exploration, continuous reevaluation, and dynamic adjustments, resulting in technical design guidelines.
• Limitations and Future Directions:
  • Limited sample diversity: participants were primarily transgender individuals with prior training experience, excluding broader cultural, racial, and early-stage training data. Future studies should include more diverse samples to address early voice training challenges.
  • Expanding the applicability of gender-affirming voice technologies: future research could explore the potential application of similar personalized goal-setting methods in other fields, such as accessible design for voice synthesis or speech therapy.
  • Further optimization of social VR "simulation environments": integrating biofeedback is recommended to allow users to monitor their voice progress in real-time and enhance immersive interactive experiences.

Through this in-depth exploration, the future development of gender-affirming voice training technologies is clearly focused on user-centered, culturally sensitive, and goal-flexible design paradigms.