首页|'Generative Model For Multi-Modality Outputs From A Single Input' in Patent Appl ication Approval Process (USPTO 20240135672)
'Generative Model For Multi-Modality Outputs From A Single Input' in Patent Appl ication Approval Process (USPTO 20240135672)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
The following quote was obtained by the news editors from the background informa tion supplied by the inventors: “A common task in digital image generation is ge nerating multiple images that depict visually similar scenes or environments, ex pressed in different modalities. For instance, some approaches for digitally ren dering a three-dimensional digital environment rely on multiple digital images t hat each express the environment in a different modality to do so, such as one i mage that depicts visual characteristics (e.g., colors) of the environment as pe rceived by a human viewer, another image that represents depth characteristics o f different points in the environment, another image that identifies segmentatio n boundaries for different objects in the environment, and so forth. Conventiona l image generation systems, however, are unable to generate these multiple modal ity images for the environment using a single generative model.”
NETL
NSTL
