Generiranje 3D geometrije pomoću alata otvorenog koda
Gespeichert in:
| Titel: | Generiranje 3D geometrije pomoću alata otvorenog koda |
|---|---|
| Autoren: | Bakšaj, Ana |
| Weitere Verfasser: | Maričić, Sven, Liverić, Lovro, Kršulja, Marko |
| Verlagsinformationen: | 2025. |
| Publikationsjahr: | 2025 |
| Schlagwörter: | speech-to-text, Python scripting, Whisper AI, Braille, FreeCAD, Python class, 3D modelling, visually impaired students |
| Beschreibung: | Ovaj rad predstavlja modularni sustav koji prevodi govor u 3D ispisive oblike Brailleovog pisma koristenjem Python skriptiranja, umjetne inteligencije Whisper i parametarskog modeliranja u FreeCAD-u. Rad sustava počinje glasovnim unosom koji se bilježi u tekstualnom obliku pomoću Whispera, AI modela otvorenog koda za automatsko prepoznavanje govora čija je glavna karakteristika visoka preciznost u različitim akustičkim uvjetima. Dobiveni tekst obrađuje Python klasa koja definira algoritam za pretvorbu znakova u Brailleov ekvivalent. Svaki znak se potom prevodi u 3D geometriju unutar FreeCAD-a čime se omogućuje generiranje fizičkih modula Brailleovog pisma. Arhitektura sustava temelji se na povezivanju Python skripte i FreeCAD makronaredbe što omogućuje dinamičko i proširivo modeliranje. Predloženo rješenje nadopunjuje suvremena istraživanja u području taktilne pedagogije i alata otvorenog koda te ističe važnost 3D modeliranja u pristupu obrazovanju i komunikaciji slijepih i slabovidnih osoba. Time se otvara prostor za nova rješenja na presjeku umjetne inteligencije, dizajna i digitalne pristupačnosti. This paper presents a modular system that translates speech into 3D printable Braille dots using Python scripting, Whisper AI, and parametric modelling in FreeCAD. The system starts with voice input that is recorded in text form using Whisper, an open source AI model for automatic speech recognition whose main feature is high accuracy in various acoustic conditions. The resulting text is processed by a Python class that defines an algorithm for converting characters into Braille equivalents. Each character is then translated into 3D geometry within FreeCAD, which enables the generation of physical Braille modules. The system architecture is based on the connection of Python script and FreeCAD macros, which enables dynamic and extensible modelling. The proposed solution complements contemporary research in the field of tactile pedagogy and open source tools, and highlights the importance of 3D modelling in accessing education and communication for blind and visually impaired people. This opens up space for new solutions at the intersection of artificial intelligence, design, and digital accessibility. |
| Publikationsart: | Master thesis |
| Dateibeschreibung: | application/pdf |
| Sprache: | Croatian |
| Zugangs-URL: | https://urn.nsk.hr/urn:nbn:hr:137:227331 |
| Rights: | URL: http://rightsstatements.org/vocab/InC/1.0/ |
| Dokumentencode: | edsair.od......4017..5648f79b22e88e34a5a6b21c7a7ed77c |
| Datenbank: | OpenAIRE |
Nájsť tento článok vo Web of Science | Abstract: | Ovaj rad predstavlja modularni sustav koji prevodi govor u 3D ispisive oblike Brailleovog pisma koristenjem Python skriptiranja, umjetne inteligencije Whisper i parametarskog modeliranja u FreeCAD-u. Rad sustava počinje glasovnim unosom koji se bilježi u tekstualnom obliku pomoću Whispera, AI modela otvorenog koda za automatsko prepoznavanje govora čija je glavna karakteristika visoka preciznost u različitim akustičkim uvjetima. Dobiveni tekst obrađuje Python klasa koja definira algoritam za pretvorbu znakova u Brailleov ekvivalent. Svaki znak se potom prevodi u 3D geometriju unutar FreeCAD-a čime se omogućuje generiranje fizičkih modula Brailleovog pisma. Arhitektura sustava temelji se na povezivanju Python skripte i FreeCAD makronaredbe što omogućuje dinamičko i proširivo modeliranje. Predloženo rješenje nadopunjuje suvremena istraživanja u području taktilne pedagogije i alata otvorenog koda te ističe važnost 3D modeliranja u pristupu obrazovanju i komunikaciji slijepih i slabovidnih osoba. Time se otvara prostor za nova rješenja na presjeku umjetne inteligencije, dizajna i digitalne pristupačnosti.<br />This paper presents a modular system that translates speech into 3D printable Braille dots using Python scripting, Whisper AI, and parametric modelling in FreeCAD. The system starts with voice input that is recorded in text form using Whisper, an open source AI model for automatic speech recognition whose main feature is high accuracy in various acoustic conditions. The resulting text is processed by a Python class that defines an algorithm for converting characters into Braille equivalents. Each character is then translated into 3D geometry within FreeCAD, which enables the generation of physical Braille modules. The system architecture is based on the connection of Python script and FreeCAD macros, which enables dynamic and extensible modelling. The proposed solution complements contemporary research in the field of tactile pedagogy and open source tools, and highlights the importance of 3D modelling in accessing education and communication for blind and visually impaired people. This opens up space for new solutions at the intersection of artificial intelligence, design, and digital accessibility. |
|---|