Li Yi

Latest

• Contrast with Reconstruct: Contrastive 3D Representation Learning Guided by Generative Pretraining
• Autoencoders as Cross-Modal Teachers: Can Pretrained 2D Image Transformers Help 3D Representation Learning?