Yi Liu,

Latest

• A generative model for vision that trains with high data efficiency and breaks text-based CAPTCHAs
• Hierarchical Compositional Feature Learning
• Teaching compositionality to CNNs