染敏光通訊介面研發
Photophone based on dye sensitized solar cell
敏光通訊介面研發 Photophone based on dye sensitized solar cell
染敏太陽能、RAVE深度學習、雷射投影
這個實驗探索了人工智慧在光學視聽裝置中的應用關係。首先，我們使用了半透明的染料敏化太陽能電池（DSSC）玻璃作為光電話，將接收到的雷射模式轉換為聲音頻率。然後，我們應用了RAVE（Real-time Audio Variational autoEncoder）模型，將雷射信號即時轉換為逼真的結果。
這個原型預計用於進一步的實驗，例如探索光學參數對訓練深度學習模型的影響，研究雷射投影模式與RAVE潛在空間（latent space）中編碼器和解碼器之間壓縮特徵之間的形態相關性，以及探索電極的圖形設計潛力，因為電極可以使用網印技術進行定制。
Dye sensitized solar cell, RAVE deep learning, laser projection
This experiment explores the relationship between AI applications in optical audiovisual installations. A semi-transparent dye-sensitized solar cell (DSSC) glass was used as a photophone to convert the received laser pattern into sound frequencies. Then, an RAVE (Real-time Audio Variational autoEncoder) model was applied to convert the laser signal into realistic results in real-time.
This prototype is expected to be used for further experiments, such as exploring the influence of optical parameters on training deep learning models, investigating the morphological correlations between laser projection patterns and the compressed features between the encoder and decoder in the latent spaces of RAVE, and exploring the potential of graphic design for the electrodes, since the electrodes can be customized using screen printing techniques.