Yuki M. Asano

Latest

• Scaling Backwards: Minimal Synthetic Pretraining?
• GeneralAD: Anomaly Detection Across Domains by Attending to Distorted Features
• Object-Centric Diffusion for Efficient Video Editing
• SIGMA: Sinkhorn-Guided Masked Video Modeling
• Learning to Count without Annotations
• Is ImageNet worth 1 video? Learning strong image encoders from 1 long unlabelled video
• VeRA: Vector-based Random Matrix Adaptation
• Self-Ordering Point Clouds
• Time Does Tell: Self-Supervised Time-Tuning of Dense Image Representations
• Self-Guided Diffusion Models
• BISCUIT: Causal Representation Learning from Binary Interactions
• The Augmented Image Prior: Distilling 1000 Classes by Extrapolating from Single Image
• Less than Few: Self-Shot Video Instance Segmentation
• VTC: Improving Video-Text Retrieval with User Comments
• CITRIS - Causal Identifiability from Temporal Intervened Sequences
• Self-Supervised Learning of Object Parts for Semantic Segmentation
• Self-supervised object detection from audio-visual correspondence
• Measuring the Interpretability of Unsupervised Representations via Quantized Reversed Probing
• Bias Out-of-the-Box: An Empirical Analysis of Intersectional Occupational Biases in Popular Generative Language Models
• Keeping Your Eye On the Ball: Trajectory Attention in Video Transformers
• PASS: An ImageNet replacement for self-supervised pretraining without humans
• Space-Time Crop & Attend: Improving Cross-modal Video Representation Learning
• On Compositions of Transformations in Contrastive Self-Supervised Learning
• Emergent inequality and business cycles in a simple behavioral macroeconomic model
• Support-set bottlenecks for video-text representation learning