<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
        <title>Teresa Pelinski and Adam Pultz Melbye – “Building and performing with an ensemble of models”</title>
        <link>https://stream.udk-berlin.de/videos/watch/e549f610-8448-4431-9af7-19ef0cdd76fa</link>
        <description>How can we then treat machine learning models as design material when building new musical instruments? Many of the tools available for embedding machine learning in musical practice have fixed model architectures, so that the practitioners’ involvement with the models is often limited to the training data. In this context, the design material appears to be the data. From a developer’s perspective, however, when training a model for a specific task, a ‘model’ does not refer to a single entity but to an ensemble instead: the architecture is trained varying the number of layers, blocks or embedding sizes (the hyperparameters), and training-specific parameters such as learning rate or dropout. For the last month we have been working on a practice-based research project involving the FAAB (feedback-actuated augmented bass), the ai-mami system (AI-models as materials interface) and a modular synthesiser. In this talk, we will discuss our process when building this performance ensemble, our insights in using machine learning models as design material, and finally, the technical infrastructure of the project. More on the event: https://www.federicovisi.com/embedding-algorithms-workshop/ Playlist: https://www.youtube.com/playlist?list=PLLO0tUhzBqFWU0YQXUrNgPyDoqwaKZxwq The Embedding Algorithms Workshop is an informal meeting of researchers and practitioners working with embedded systems in the fields of musical instrument design, wearable computing, interaction design, and performing arts. During the workshop, the participants will show and discuss: designing physical objects (instruments, wearables, and more) that incorporate some algorithmic process that is central to their function or behaviour; designing and implementing embedded algorithms; using/performing/practising with objects with embedded algorithms; and more.</description>
        <lastBuildDate>Fri, 20 Nov 2026 23:39:04 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>PeerTube - https://stream.udk-berlin.de</generator>
        <image>
            <title>Teresa Pelinski and Adam Pultz Melbye – “Building and performing with an ensemble of models”</title>
            <url>https://stream.udk-berlin.de/lazy-static/avatars/017ba0a6-fb38-49e1-bd60-32d14974dca6.png</url>
            <link>https://stream.udk-berlin.de/videos/watch/e549f610-8448-4431-9af7-19ef0cdd76fa</link>
        </image>
        <copyright>All rights reserved, unless otherwise specified in the terms specified at https://stream.udk-berlin.de/about and potential licenses granted by each content's rightholder.</copyright>
        <atom:link href="https://stream.udk-berlin.de/feeds/video-comments.xml?videoId=e549f610-8448-4431-9af7-19ef0cdd76fa" rel="self" type="application/rss+xml"/>
    </channel>
</rss>