MULTIMODAL GENERATIVE MODELS FOR MOBILE-FIRST CONSUMER APPS AN END-TO-END ARCHITECTURE
Fuente:
Zenodo
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Recurso digital |
| Pubblicazione: |
Zenodo
2024
|
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866901542086377472 |
|---|---|
| author | Faruk Baran Öncel |
| author_facet | Faruk Baran Öncel |
| contents | <p><a href="https://ijetrm.com/issues/files/Jun-2024-08-1749368531-MAR202432.pdf" target="_blank" rel="noopener">Generative AI</a> has made it possible for consumers to find interactive, place-based and various forms of content in<br>mobile device applications. This paper explains how text, image, audio and video can be handled in mobile<br>consumer applications using generative technology. Because mobile devices are being used more for computing, we<br>must make sure these powerful models work well when resources are scarce. We focus on conducting work for<br>mobile devices, where you will find techniques on picking models, training, handling data efficiently and knowledge<br>compression. In addition, we suggest mixing local and cloud methods to manage big data quickly and efficiently.<br>The paper explains techniques for uniting the outcomes of various models while making certain to ensure user<br>privacy. Examples are found in the use of live content generation, personalized recommendations, conversational<br>robots and augmented reality. We offer developers in the field practical advice, specifications and useful guidelines<br>to assist in creating multimodal systems at the mobile edge. As a result, this work can be used to design and build<br>the next generation of smart AI-driven mobile programs.</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_15617984 |
| institution | Zenodo |
| language | |
| publishDate | 2024 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | MULTIMODAL GENERATIVE MODELS FOR MOBILE-FIRST CONSUMER APPS AN END-TO-END ARCHITECTURE Faruk Baran Öncel <p><a href="https://ijetrm.com/issues/files/Jun-2024-08-1749368531-MAR202432.pdf" target="_blank" rel="noopener">Generative AI</a> has made it possible for consumers to find interactive, place-based and various forms of content in<br>mobile device applications. This paper explains how text, image, audio and video can be handled in mobile<br>consumer applications using generative technology. Because mobile devices are being used more for computing, we<br>must make sure these powerful models work well when resources are scarce. We focus on conducting work for<br>mobile devices, where you will find techniques on picking models, training, handling data efficiently and knowledge<br>compression. In addition, we suggest mixing local and cloud methods to manage big data quickly and efficiently.<br>The paper explains techniques for uniting the outcomes of various models while making certain to ensure user<br>privacy. Examples are found in the use of live content generation, personalized recommendations, conversational<br>robots and augmented reality. We offer developers in the field practical advice, specifications and useful guidelines<br>to assist in creating multimodal systems at the mobile edge. As a result, this work can be used to design and build<br>the next generation of smart AI-driven mobile programs.</p> |
| title | MULTIMODAL GENERATIVE MODELS FOR MOBILE-FIRST CONSUMER APPS AN END-TO-END ARCHITECTURE |
| url | https://doi.org/10.5281/zenodo.15617984 |