情緒現形

本作品以語音情緒分析為基礎，探討人在敘述事件時所蘊含的多重情緒結構。人們往往將一段經驗簡化為單一情緒，然而語言之中實則隱含著複雜且交織的情緒層次。作品透過錄音擷取使用者的語音輸入，並即時傳送至雲端語音情緒分析系統，依據語音特徵與口語用詞辨識出多種情緒成分，並轉化為可視化參數。

在視覺呈現上，情緒數據被傳輸至 TouchDesigner 作為觸發依據，不同情緒對應不同視覺表現，並依其數值影響畫面的動態速度與狀態。各種情緒以不同形式出現並相互疊加，形成具層次與流動感的視覺結構。這些視覺不僅反映單一情緒，而是呈現多種情緒同時存在的狀態，使觀者得以直觀地理解情緒之間的關係與變化。

透過「說」與「看」的互動過程，作品引導觀者重新理解自身經驗，意識到情緒並非單一且固定，而是流動且多層的存在。藉由將不可見的情緒轉譯為可感知的視覺形式，作品試圖打破既有的情緒認知框架，開啟對自我感受更細緻的覺察。

Revealed

This project is based on speech emotion analysis, exploring the layered emotions embedded in how people narrate their experiences. While events are often simplified into a single emotion, language contains complex and overlapping emotional layers. The system captures voice input and transmits it in real time to a cloud-based model, which identifies multiple emotions based on vocal features and spoken language, and converts them into visual parameters.

These data are sent to TouchDesigner, where different emotions trigger corresponding visual forms. Their intensity influences the speed and behavior of the visuals, which overlap to create a dynamic and layered structure. Rather than representing a single emotion, the visuals reveal the coexistence of multiple emotional states.

Through the process of “speaking” and “seeing,” the project invites viewers to recognize emotion as fluid and multi-layered, offering a new way to understand their inner experiences.