DC-AE-f-32

Emerging

1papers using it

2025first seen

The 'DC-AE-f-32' dataset/benchmark is used to evaluate the performance of high-dimensional autoencoders in terms of generation quality and reconstruction fidelity, particularly focusing on the impact of high-frequency latent components.

🔎 Find this dataset

Papers using DC-AE-f-32 (1)

Toward Diffusible High-Dimensional Latent Spaces: A Frequency Perspective2025