Latest

• Compact Language Models via Pruning and Knowledge Distillation
• MaskLLM: Learnable Semi-Structured Sparsity for Large Language Models
• COIN: Control-Inpainting Diffusion Prior for Human and Camera Motion Estimation
• LITA: Language Instructed Temporal-localization Assistant
• Flextron: Many-in-One Flexible Large Language Model
• AM-RADIO: Agglomerative Model - Reduce All Domains Into One
• VILA: On pretraining for vision language models
• FasterViT: Fast Vision Transformers with Hierarchical Attention
• PACE: Human and Camera Motion Estimation from in-the-wild Videos
• RANA: Relightable and Articulated Neural Avatars
• Global Context Vision Transformers
• Global Vision Transformer Pruning with Hessian-Aware Saliency
• Heterogeneous Continual Learning
• Recurrence without Recurrence: Stable Video Landmark Detection with Deep Equilibrium Models
• Structural Pruning via Latency-Saliency Knapsack
• Towards Annotation-efficient Segmentation via Image-to-image Translation
• LANA: Latency Aware Network Acceleration
• Privacy Vulnerability of Split Computing to Data-Free Model Inversion Attacks
• A-ViT: Adaptive Tokens for Efficient Vision Transformer
• GLAMR: Global Occlusion-Aware Human Mesh Recovery with Dynamic Cameras
• GradViT: Gradient Inversion of Vision Transformers
• When to Prune? A Policy towards Early Structural Pruning
• DRaCoN--Differentiable Rasterization Conditioned Neural Radiance Fields for Articulated Avatars
• Do Gradient Inversion Attacks Make Federated Learning Unsafe?
• KAMA: 3D Keypoint Aware Body Mesh Articulation
• NViT: Vision Transformer Compression and Parameter Redistribution
• Optimizing Selective Protection for CNN Resilience
• Adversarial Motion Modelling Helps Semi-Supervised Hand Pose Estimation
• DexYCB: A Benchmark for Capturing Hand Grasping of Objects
• Optimal Quantization Using Scaled Codebook
• See through Gradients: Image Batch Recovery via GradInversion
• Weakly-Supervised 3D Human Pose Learning via Multi-view Images in the Wild