mlsanityData to 7 Oct 2026

Dynamic Filter Networks

arXiv 1605.09673 ↗First posted 31 May 2016cs.LG cs.CVcited by 318

Not on the map. Why?

Abstract (arXiv)

In a traditional convolutional layer, the learned filters stay fixed after training. In contrast, we introduce a new framework, the Dynamic Filter Network, where filters are generated dynamically conditioned on an input. We show that this architecture is a powerful one, with increased flexibility thanks to its adaptive nature, yet without an excessive increase in the number of model parameters. A wide variety of filtering operations can be learned this way, including local spatial transformations, but also others like selective (de)blurring or adaptive feature extraction. Moreover, multiple such layers can be combined, e.g. in a recurrent architecture. We demonstrate the effectiveness of the dynamic filter network on the tasks of video and stereo prediction, and reach state-of-the-art performance on the moving MNIST dataset with a much smaller model. By visualizing the learned filters, we illustrate that the network has picked up flow information by only looking at unlabelled training data. This suggests that the network can be used to pretrain networks for various supervised tasks in an unsupervised way, like optical flow and depth estimation.

Key facts. Cited by 318 papers in the map (data through 7 Oct 2026). First posted on arXiv 31 May 2016.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 114; 2022: 64; 2023: 55; 2024: 27; 2025: 34; 2026: 24.

318 papers.

Sort:

By area:

  1. 2021Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutionscited by 1,960
  2. 2021Domain Generalization: A Surveycited by 760
  3. 2023Scaling up GANs for Text-to-Image Synthesiscited by 492
  4. 2021Linear Transformers Are Secretly Fast Weight Programmerscited by 471
  5. 2021Conditional DETR for Fast Training Convergencecited by 422
  6. 2021Dynamic Neural Networks: A Surveycited by 405
  7. 2021XCiT: Cross-Covariance Image Transformerscited by 335
  8. 2022Mask3D: Mask Transformer for 3D Semantic Instance Segmentationcited by 299
  9. 2021PAConv: Position Adaptive Convolution with Dynamic Kernel Assembling on Point Cloudscited by 211
  10. 2021K-Net: Towards Unified Image Segmentationcited by 211
  11. 2021PredRNN: A Recurrent Neural Network for Spatiotemporal Predictive Learningcited by 181
  12. 2022SimVP: Simpler yet Better Video Predictioncited by 181
38 more
  1. 2021Fast Convergence of DETR with Spatially Modulated Co-Attentioncited by 176
  2. 2021Deep Parametric Continuous Convolutional Neural Networkscited by 161
  3. 2021Instances as Queriescited by 157
  4. 2023SparseBEV: High-Performance Sparse 3D Object Detection from Multi-Camera Videoscited by 149
  5. 2021Efficient DETR: Improving End-to-End Object Detector with Dense Priorcited by 128
  6. 2022HorNet: Efficient High-Order Spatial Interactions with Recursive Gated Convolutionscited by 103
  7. 2021Involution: Inverting the Inherence of Convolution for Visual Recognitioncited by 99
  8. 2021CondLaneNet: a Top-to-down Lane Detection Framework Based on Conditional Convolutioncited by 95
  9. 2021Lite-HRNet: A Lightweight High-Resolution Networkcited by 91
  10. 2021LAPAR: Linearly-Assembled Pixel-Adaptive Regression Network for Single Image Super-Resolution and Beyondcited by 88
  11. 2021FitVid: Overfitting in Pixel-Level Video Predictioncited by 86
  12. 2021Iterative Filter Adaptive Network for Single Image Defocus Deblurringcited by 85
  13. 2021Going Beyond Linear Transformers with Recurrent Fast Weight Programmerscited by 84
  14. 2022Temporal Attention Unit: Towards Efficient Spatiotemporal Predictive Learningcited by 84
  15. 2021Crossover Learning for Fast Online Video Instance Segmentationcited by 83
  16. 2021Cross-domain Few-shot Learning with Task-specific Adapterscited by 79
  17. 2022AdaMixer: A Fast-Converging Query-Based Object Detectorcited by 78
  18. 2021Greedy Hierarchical Variational Autoencoders for Large-Scale Video Predictioncited by 75
  19. 2021Open-World Entity Segmentationcited by 74
  20. 2024DiM: Diffusion Mamba for Efficient High-Resolution Image Synthesiscited by 72
  21. 2022Dynamic Spatial Propagation Network for Depth Completioncited by 70
  22. 2022SwinTextSpotter: Scene Text Spotting via Better Synergy between Text Detection and Text Recognitioncited by 70
  23. 2021SOLQ: Segmenting Objects by Learning Queriescited by 68
  24. 2021Predicting Video with VQVAEcited by 61
  25. 2021On the Connection between Local Attention and Dynamic Depth-wise Convolutioncited by 61
  26. 2021GAN-Supervised Dense Visual Alignmentcited by 61
  27. 2023Fully Sparse 3D Occupancy Predictioncited by 61
  28. 2021Adaptive Graph Convolution for Point Cloud Analysiscited by 60
  29. 2022CMT-DeepLab: Clustering Mask Transformers for Panoptic Segmentationcited by 60
  30. 2022Towards Spatio-Temporal Aware Traffic Time Series Forecasting--Full Versioncited by 59
  31. 2023Learning to Upsample by Learning to Samplecited by 57
  32. 2022Omni-Dimensional Dynamic Convolutioncited by 55
  33. 2021MotionRNN: A Flexible Model for Video Prediction with Spacetime-Varying Motionscited by 54
  34. 2021Convolutional Dynamic Alignment Networks for Interpretable Classificationscited by 54
  35. 2021Decoupled Dynamic Filter Networkscited by 52
  36. 2023Adaptive Rotated Convolution for Rotated Object Detectioncited by 51
  37. 2022SepLUT: Separable Image-adaptive Lookup Tables for Real-time Image Enhancementcited by 50
  38. 2023LRRU: Long-short Range Recurrent Updating Networks for Depth Completioncited by 48

50 most cited of 318 All 318 (JSON)