Tian Han

Researcher and Assistant Professor in Stevens Institute of Tech. Works on statistical generative modeling, deep learning and explainable AI
PhD, UCLA

[Tian Han @TianHan10]

Check this out - our new ICML24 paper about adversarial energy-based model. The EBM can be better and more efficiently learned with diffusion!

[Cong Geng @conggeng94]

I'll be attending #ICML2024 to present our paper "Improving Adversarial Energy-Based Model via Diffusion Process" in poster session 4(Wed 24 Jul 1:30 p.m.), if you are also interested in EBM, do stop by and chat with me!

preprint: https://t.co/jTszOrnPyP

[Tian Han @TianHan10]

@KL_Div @yszhang170 @PengShichong @Moazeni_Alireza @NeurIPSConf Congrats, Ke!

[Tian Han @TianHan10]

Excited to share our oral presented work in NeurIPS2022 "Adaptive Multi-stage Density Ratio Estimation for Learning Latent Space Energy-based Model": https://t.co/tLWSlAU26p. Latent EBM is learned through multiple stages of density ratios via NCE (no mcmc). #NeurIPS2022 #ebms

[Tian Han @TianHan10]

@FengShi25242729 😁

[Tian Han @TianHan10]

Glad to share our recent ECCV work on learning multi-layer latent variable model via short run MCMC. The model is also deeply rooted in our previous work on Alternating Back-Propagation (ABP) https://t.co/K8oNMPAPgX. Both works studied the MLE learning of the generator model.

[Erik Nijkamp @erik_nijkamp]

Grateful to have our work accepted in ECCV 2020 with amazing co-authors @TianHan10, @bo_pang0, et al.,
"Learning multi-layer latent variable model via variational optimization of short run MCMC for approximate inference":
https://t.co/ENVrrjdz9R
#ECCV2020 #mcmc

[Tian Han @TianHan10]

@sfrei_ @hugo_larochelle Oh, me too. This is quite strange!

[Tian Han @TianHan10]

@ChunyuanLi Nice work!

[Tian Han @TianHan10]

@sfrei_ @_YuanCao_ @QuanquanGu Great work, Spencer!

[Tian Han @TianHan10]

@jaschasd Wow. It's an inspiring paper indeed, Jascha. I would also recommend our recent work which also considers energy based model in the latent space. It improves both the image and text modeling (https://t.co/Tq3FVsuKz0).

[Tian Han @TianHan10]

@fregu856 Thanks for your interest! Perhaps we could collaborate on some topics in the future..

[Tian Han @TianHan10]

Thrilled to share our recent work: Learning Latent Space Energy-Based Prior Model (https://t.co/ZmABekOSLa). Joint work with Bo Pang and Erik Nijkamp @erik_nijkamp. We build EBM on the latent space of the generator and learn it through prior and posterior Langevin Dynamics.