Unverified50% confidenceFactExact time
2015年Gatys等人发表的《A Neural Algorithm of Artistic Style》开创了深度学习时代的图像艺术化范式，利用预训练VGG网络的特征并通过Gram矩阵衡量风格相似度
1
Sources
50%
Confidence
Long-term
Relevance
7/17/2026
First Seen
Sources
Related Claims
Verified2015年德国图宾根大学的Leon Gatys等人发表论文《A Neural Algorithm of Artistic Style》，首次证明卷积神经网络能够分离并重组图像的内容与风格74% similarUnverifiedAdobe Photoshop的内容识别填充和神经滤镜功能均基于深度学习模型72% similarUnverified2015年Gatys等人提出神经风格迁移，通过卷积神经网络的Gram矩阵来匹配内容图与风格图的特征分布70% similarUnverifiedConvolutional Neural Networks automatically learn hierarchical feature representations of images through convolutional layers, from low-level edges and textures to high-level object parts and semantic concepts.69% similarUnverified2015年Gatys等人提出的神经风格迁移算法可将一幅画的艺术风格迁移到另一幅图像上，这一思想随后被迁移到NLP领域68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/544835API
curl https://kongchang.com/api/v1/knowledge/claims/544835MCP
get_claim(id=544835)