Fetching the paper…

AR-1-to-3: Single Image to Consistent 3D Object Generation via Next-View Prediction · Around