Skip To Main Content
Toshimitsu Uesaka on Muck Rack

Toshimitsu Uesaka

As seen in: arXiv
Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work. Claim your profile

Get in touch with Toshimitsu

Contact Toshimitsu, search articles and posts on X, monitor coverage, and track replies from one place.

Learn more about Muck Rack

Actions

Share this page

Is this you?

As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile

Articles

[2410.14758] Mitigating Embedding Collapse in Diffusion Models for Categorical Data

|
By Yuhta Takida, Naoki Murata, Toshimitsu Uesaka, Chieh-Hsin Lai
| arXiv
Open access is only possible with YOUR support. Give to arXiv this week to help keep science open for all.
Open in Who Shared View Duplicates (631) Issue with article?

Tutorial: Step-by-Step Guide to SAN-ifying Your GAN Models for Improved Performance – Sony AI

|
By Yuki Mitsufuji, Takashi Shibuya, Yuhta Takida, Toshimitsu Uesaka
| ai.sony
This tutorial will give a way to implement Slicing Adversarial Networks (SAN), a universal approach, just like spectral normalization, to enhance GAN [1] performance. SAN serves as a drop-in replacement for GAN, since almost any GAN can be easily converted to SAN with a couple of small changes to a discriminator. More specifically, all you need for converting GANs into the SAN counterparts is modifying the last linear layer of a discriminator and the discriminator objective function.
Open in Who Shared Issue with article?
See all 2 articles