Only Text Isixhobo Sokwakha Uyilo Lwangaphakathi Ngokuzenzekelayo ngu-Zichun Shao

Uyilo olubizwa ngokuba yi-Only Text Isixhobo Sokwakha Uyilo Lwangaphakathi Ngokuzenzekelayo olwenziwe ngu-Zichun Shao ngumwini obekekileyo webhaso elinobukhazikhazi le-2023 Ibhaso Lentsimbi le-A' Design Award Kudidi Lokhuphiswano lwe-Uyilo oluVelisayo, olwe-Algorithimu, olweParametrikhi kunye noluNcediswa yi-AI.

IntsimbiOluvelisayo· 2023

Only Text Isixhobo Sokwakha Uyilo Lwangaphakathi Ngokuzenzekelayo

ngu Zichun Shao

Umxhasi: Zichun Shao

Umwini we-Ibhaso Lentsimbi le-A' Design Award Kudidi Lwe-Oluvelisayo

44,535

Umwini Webhaso Oqinisekisiweyo

Qinisekisa Isimo Sebhaso se-#150725

Jonga kwi-A' Design Award Uxwebhu Lweendaba Ibheji Yembasa

Imibala

Uyilo

Malunga ne-Only Text Isixhobo Sokwakha Uyilo Lwangaphakathi Ngokuzenzekelayo

I-Only Text sisixhobo sokwakha uyilo ngokuzenzekelayo esidala imifanekiso yoyilo lwangaphakathi ngokusuka kwiinkcazelo ezibhaliweyo. Isixhobo sisekelwe kumodeli yokusasazeka (diffusion model) eqeqeshelwe uyilo lwangaphakathi lobungcali. Ngaphezu kwe-20,000 zomfanekiso wangaphakathi zaqokelelwa kwaza zaphawulwa ngaboyili ukwenza isethi yedata yoqeqesho. I-Only Text inako ukuvelisa imboniso yangaphakathi ngeemizuzwana embalwa, ngesitayela sokuhombisa esikhethiweyo neempawu zesithuba. Iinkcukacha zitshintshwa ngokuhlela umbhalo nokuvelisa kwakhona. Le ndlela isusa umsebenzi wokwenza imodeli nokuvelisa ngesandla, ngoko iinketho ezininzi zinganatho, kwaneminikelo ekulishumi.

Ukuqatshelwa

Malunga ne-Ibhaso Lentsimbi le-A' Design Award

Only Text Isixhobo Sokwakha Uyilo Lwangaphakathi Ngokuzenzekelayo ngumphumeleli weIbhaso Lentsimbi le-A' Design Award, ukuqatshelwa okuhlonitshwayo okunikwa imiyilo elungileyo eyanelisa iimfuno zobungcali neemveliso ezilindelekileyo kwimveliso ezicwangciswe kakuhle. Abaphumeleleyo beNtsimbi babonisa ubuchule obukhulu, ubuchule obuphezulu, ukuqonda kwengcali namandla okuyila abayili babo. Only Text Isixhobo Sokwakha Uyilo Lwangaphakathi Ngokuzenzekelayo iyathenjwa ukuba ibe yimveliso exatyiswa kakhulu, edunyiswayo nethandwayo esebenzayo neyinto entsha. Lo myilo uphumeleleyo weIbhaso Lentsimbi le-A' Design Award udibanisa iindlela ezilungileyo kakhulu zemboleko kwaye unezimpawu zobuchwephesha ezifanelekileyo ezidityaniswe nolwazi netalente. Only Text Isixhobo Sokwakha Uyilo Lwangaphakathi Ngokuzenzekelayo iyahlonelwa ngenxa yenkcukacha yayo kwaye yacwangciselwa ukunika uphuculo lomgangatho wobomi olunceda ukwenza ihlabathi libe yindawo engcono.

Emva Koyilo

Ikhithi Yeendaba

Iziqinisekiso, amalungu eqela kunye neenkcukacha zobugcisa ezinikezelwe ngumyili.

Iimpawu Ezizodwa: The design was done by artificial intelligence (AI) in collaboration with designers. This project has trained an AI for professional interior design generation, which can generate interior designs with different decoration styles according to the input text. It only takes a few seconds for AI to generate a new interior design, and it can quickly generate a large number of designs for communication with users, thereby improving design efficiency.
Imilinganiselo nePakishini: The project uses diffusion model technology, using artificial intelligence to cooperate with designers to generate designs of different decorative styles. We retrained a professional interior design AI to quickly create designs with varying decoration styles.
Iithegi: Artificial intelligence, Human-computer collaboration, Interior design, Generative design, Text generative design, Efficient design, Rapid modification, Controllable design, Decoration style control
Ubuchwepheshe Bokuvelisa: Ordinary AI cannot generate interior design according to specific "decoration style" and "space attributes," which is a fatal flaw for interior design. To this end, this project collected more than 20,000 interior design renderings, labeled them to make a data set and gave the data set to AI for learning, thus creating an AI suitable for interior design. The AI can generate interior renderings specifying "decorative style" and "space attributes" by inputting text.
Umngeni Woyilo: The challenge of this design is to build a professional interior design AI that generates designs of high quality and correct decor. To this end, this project asked professional designers to collect and label tens of thousands of interior design images and then handed over the filtered data to AI for training. Make AI understand what is "decorative style," "spatial attributes," and "aesthetic attributes" and generate an interior design with a specified decorative style.
Uphando: Research Background: Interior design communication is complex, and obtaining designs and communicating with customers quickly is necessary.
Methods: This project trained an AI to generate interior designs professionally and used the AI to collaborate with humans to create interior designs in various decorative styles from the text.
Results: Using the AI trained in this project, designers can generate an interior design rendering in just a few seconds.
Insights: This project approach rapidly generates large batches of interior designs for specified decor styles and spatial properties and allows new designs to be generated by modifying text. This method eliminates the need for manual modeling and rendering, which improves design efficiency.
Inkuthazo: It is difficult for users to imagine the scheme described by the designer in words. Designers need to turn thinking into design quickly. With the development of artificial intelligence (AI), designers can generate designs directly from text. Specifically, this project trained an AI to create interior design, which can generate interior design only by inputting text, improving communication efficiency between designers and users.
Ubude Beprojekthi: The project is located in Guangzhou, Guangdong Province, China, and was designed in 2023.
Ukuhamba Komsebenzi: 1. Professional designers collect interior design images and mark and screen them.
2. Use the filtered data to train a professional interior design AI.
3. Designers use AI to design.
4. The designer inputs the design text description into the AI to specify the decoration style.
5. AI generates the interior design of the specified decoration style.
6. Modify the unsatisfactory details in the design through text description, and regenerate the design.
7. After adjusting the text description and generating it multiple times, a good design is obtained.