Description
Book SynopsisThe 39-volume set, comprising the LNCS books 13661 until 13699, constitutes the refereed proceedings of the 17th European Conference on Computer Vision, ECCV 2022, held in Tel Aviv, Israel, during October 23–27, 2022.
The 1645 papers presented in these proceedings were carefully reviewed and selected from a total of 5804 submissions. The papers deal with topics such as computer vision; machine learning; deep neural networks; reinforcement learning; object recognition; image classification; image processing; object detection; semantic segmentation; human pose estimation; 3d reconstruction; stereo vision; computational photography; neural networks; image coding; image reconstruction; object recognition; motion estimation.
Table of ContentsA Simple Approach and Benchmark for 21,000-Category Object Detection.- Knowledge Condensation Distillation.- Reducing Information Loss for Spiking Neural Networks.- Masked Generative Distillation.- Fine-Grained Data Distribution Alignment for Post-Training Quantization.- Learning with Recoverable Forgetting.- Efficient One Pass Self-Distillation with Zipf’s Label Smoothing.- Prune Your Model before Distill It.- Deep Partial Updating: Towards Communication Efficient Updating for On-Device Inference.- Patch Similarity Aware Data-Free Quantization for Vision Transformers.- L3: Accelerator-Friendly Lossless Image Format for High-Resolution, High-Throughput DNN Training.- Streaming Multiscale Deep Equilibrium Models.- Symmetry Regularization and Saturating Nonlinearity for Robust Quantization.- SP-Net: Slowly Progressing Dynamic Inference Networks.- Equivariance and Invariance Inductive Bias for Learning from Insufficient Data.- Mixed-Precision Neural Network Quantization via Learned Layer-Wise Importance.- Event Neural Networks.- EdgeViTs: Competing Light-Weight CNNs on Mobile Devices with Vision Transformers.- PalQuant: Accelerating High-Precision Networks on Low-Precision Accelerators.- Disentangled Differentiable Network Pruning.- IDa-Det: An Information Discrepancy-Aware Distillation for 1-Bit Detectors.- Learning to Weight Samples for Dynamic Early-Exiting Networks.- AdaBin: Improving Binary Neural Networks with Adaptive Binary Sets.- Adaptive Token Sampling for Efficient Vision Transformers.- Weight Fixing Networks.- Self-Slimmed Vision Transformer.- Switchable Online Knowledge Distillation.- ℓ∞-Robustness and Beyond: Unleashing Efficient Adversarial Training.- Multi-Granularity Pruning for Model Acceleration on Mobile Devices.- Deep Ensemble Learning by Diverse Knowledge Distillation for Fine-Grained Object Classification.- Helpful or Harmful: Inter-Task Association in Continual Learning.- Towards Accurate Binary Neural Networks via Modeling Contextual Dependencies.- SPIN: An Empirical Evaluation on Sharing Parameters of Isotropic Networks.- Ensemble Knowledge Guided Sub-network Search and Fine-Tuning for Filter Pruning.- Network Binarization via Contrastive Learning.- Lipschitz Continuity Retained Binary Neural Network.- SPViT: Enabling Faster Vision Transformers via Latency-Aware Soft Token Pruning.- Soft Masking for Cost-Constrained Channel Pruning.- Non-uniform Step Size Quantization for Accurate Post-Training Quantization.- SuperTickets: Drawing Task-Agnostic Lottery Tickets from Supernets via Jointly Architecture Searching and Parameter Pruning.- Meta-GF: Training Dynamic-Depth Neural Networks Harmoniously.- Towards Ultra Low Latency Spiking Neural Networks for Vision and Sequential Tasks Using Temporal Pruning.- Towards Accurate Network Quantization with Equivalent Smooth Regularizer.
