[HN Gopher] StarVector: Generating Scalable Vector Graphics Code...
       ___________________________________________________________________
        
       StarVector: Generating Scalable Vector Graphics Code from Images
       and Text
        
       Author : lnyan
       Score  : 62 points
       Date   : 2025-03-22 02:25 UTC (4 days ago)
        
 (HTM) web link (starvector.github.io)
 (TXT) w3m dump (starvector.github.io)
        
       | iFire wrote:
       | Horray. Code is released
        
       | peter_d_sherman wrote:
       | This would be absolutely _GREAT_ for generating icons for
       | applications!
       | 
       | (Also would make a great SaaS... for $X/month ($9.95, $19.95,
       | ??.??) generate unlimited icons...)
       | 
       | Congrats to the team for their pioneering hard work in this
       | nascent area of LLM/Transformer research!
       | 
       | Well done!
        
         | smahs wrote:
         | No it won't (most likely). VTracer (which the authors compare
         | with) is fast, runs in browser via wasm, consumes way less
         | resources and can even convert natural images very decently.
         | But the model seems cool for the usecase of prompt to logo or
         | icon (over my current workflow of getting a jpg from flux and
         | passing it through VTracer). I hope someone over at llama.cpp
         | notices this (at least for the text-to-svg usecase, if not
         | multimodal).
        
       | aikinai wrote:
       | I've been waiting for someone to make something like this!
       | Perfect.
        
       | rahimnathwani wrote:
       | Note: StarVector models will not work for natural images or
       | illustrations, as they have not been trained on those images.
       | They excel in vectorizing icons, logotypes, technical diagrams,
       | graphs, and charts.
        
       | jasonjmcghee wrote:
       | Did anyone else notice - the molecule it generated did not match
       | the source image.
       | 
       | Seems like this could be incredibly valuable, but I'd argue there
       | needs to be validation steps in place to confirm it's actually
       | generating the right thing, for the case of image -> vector
       | generation.
        
       ___________________________________________________________________
       (page generated 2025-03-26 23:02 UTC)