`

Timezone: »

 
Poster
Further Analysis of Outlier Detection with Deep Generative Models
Ziyu Wang · Bin Dai · David P Wipf · Jun Zhu

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1542
in Poster Session 6 »

The recent, counter-intuitive discovery that deep generative models (DGMs) can frequently assign a higher likelihood to outliers has implications for both outlier detection applications as well as our overall understanding of generative modeling. In this work, we present a possible explanation for this phenomenon, starting from the observation that a model's typical set and high-density region may not conincide. From this vantage point we propose a novel outlier test, the empirical success of which suggests that the failure of existing likelihood-based outlier tests does not necessarily imply that the corresponding generative model is uncalibrated. We also conduct additional experiments to help disentangle the impact of low-level texture versus high-level semantics in differentiating outliers. In aggregate, these results suggest that modifications to the standard evaluation practices and benchmarks commonly applied in the literature are needed.

[ Paper ]

Author Information

Ziyu Wang (Tsinghua University)
Bin Dai (Samsung Research China - Beijing)
David P Wipf (AWS)
Jun Zhu (Tsinghua University)

More from the Same Authors