mlsanityData to 7 Oct 2026

Recurrence without Recurrence: Stable Video Landmark Detection with Deep Equilibrium Models

arXiv 2304.00600 ↗First posted 2 Apr 2023cs.CV cs.LGcited by 26

Not on the map. Why?

Abstract (arXiv)

Cascaded computation, whereby predictions are recurrently refined over several stages, has been a persistent theme throughout the development of landmark detection models. In this work, we show that the recently proposed Deep Equilibrium Model (DEQ) can be naturally adapted to this form of computation. Our Landmark DEQ (LDEQ) achieves state-of-the-art performance on the challenging WFLW facial landmark dataset, reaching 3.92 NME with fewer parameters and a training memory cost of O(1) in the number of recurrent modules. Furthermore, we show that DEQs are particularly suited for landmark detection in videos. In this setting, it is typical to train on still images due to the lack of labelled videos. This can lead to a ``flickering'' effect at inference time on video, whereby a model can rapidly oscillate between different plausible solutions across consecutive frames. By rephrasing DEQs as a constrained optimization, we emulate recurrence at inference time, despite not having access to temporal data at training time. This Recurrence without Recurrence (RwR) paradigm helps in reducing landmark flicker, which we demonstrate by introducing a new metric, normalized mean flicker (NMF), and contributing a new facial landmark video dataset (WFLW-V) targeting landmark uncertainty. On the WFLW-V hard subset made up of 500 videos, our LDEQ with RwR improves the NME and NMF by 10 and 13% respectively, compared to the strongest previously published model using a hand-tuned conventional filter.

Key facts. Cited by 26 papers in the map (data through 7 Oct 2026). Builds on 46 of its 89 references that are in the map. First posted on arXiv 2 Apr 2023.

Builds on

46 of 89 references.

References · 46

  1. 2018Neural Ordinary Differential Equationscited by 3,733
  2. 2019Deep High-Resolution Representation Learning for Human Pose Estimationcited by 1,339
  3. 2016Stacked Hourglass Networks for Human Pose Estimationcited by 913
  4. 2016Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networkscited by 895
  5. 2019Deep Equilibrium Modelscited by 603
  6. 2017OptNet: Differentiable Optimization as a Layer in Neural Networkscited by 588
  7. 2019Augmented Neural ODEscited by 396
  8. 2015WIDER FACE: A Face Detection Benchmarkcited by 297
  9. 2017"Zero-Shot" Super-Resolution using Deep Internal Learningcited by 285
  10. 2018Deep Face Recognition: A Surveycited by 275
  11. 2018Deep Facial Expression Recognition: A Surveycited by 269
  12. 2017VEEGAN: Reducing Mode Collapse in GANs using Implicit Variational Learningcited by 199
34 more
  1. 2021General Facial Representation Learning in a Visual-Linguistic Mannercited by 193
  2. 2020Adaptive Risk Minimization: Learning to Adapt to Domain Shiftcited by 183
  3. 2018Recasting Gradient-Based Meta-Learning as Hierarchical Bayescited by 180
  4. 2019SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solvercited by 173
  5. 2021Hand-Object Contact Consistency Reasoning for Human Grasps Generationcited by 160
  6. 2020Multiscale Deep Equilibrium Modelscited by 149
  7. 2021HRFormer: High-Resolution Transformer for Dense Predictioncited by 149
  8. 2018Look at Boundary: A Boundary-Aware Face Alignment Algorithmcited by 137
  9. 2019Adaptive Wing Loss for Robust Face Alignment via Heatmap Regressioncited by 127
  10. 2020Self-Supervised Policy Adaptation during Deploymentcited by 121
  11. 2021A-SDF: Learning Disentangled Signed Distance Functions for Articulated Shape Representationcited by 113
  12. 2017Wing Loss for Robust Facial Landmark Localisation with Convolutional Neural Networkscited by 108
  13. 2020Monotone operator equilibrium networkscited by 100
  14. 2020LUVLi Face Alignment: Estimating Landmarks' Location, Uncertainty, and Visibility Likelihoodcited by 87
  15. 2021On Training Implicit Modelscited by 73
  16. 2021Stabilizing Equilibrium Models by Jacobian Regularizationcited by 71
  17. 2018Facial Landmark Detection: a Literature Surveycited by 63
  18. 2022Deep Equilibrium Optical Flow Estimationcited by 63
  19. 2017Human Pose Regression by Combining Indirect Part Detection and Contextual Informationcited by 52
  20. 2020Structured Landmark Detection via Topology-Adapting Deep Graph Learningcited by 50
  21. 2020Lipschitz Bounded Equilibrium Networkscited by 49
  22. 2022Sparse Local Patch Transformer for Robust Face Alignment and Landmarks Inherent Relation Learningcited by 47
  23. 2019DeCaFA: Deep Convolutional Cascade for Face Alignment In The Wildcited by 46
  24. 2019Aggregation via Separation: Boosting Facial Landmark Detector with Semi-Supervised Style Translationcited by 33
  25. 2021Test-Time Personalization with a Transformer for Human Pose Estimationcited by 33
  26. 2021ADNet: Leveraging Error-Bias Towards Normal Direction in Face Alignmentcited by 33
  27. 2021Fast Facial Landmark Detection and Applications: A Surveycited by 18
  28. 2017Facial Expression Recognition Using Enhanced Deep 3D Convolutional Neural Networkscited by 17
  29. 2020Robust Face Alignment by Multi-order High-precision Hourglass Networkcited by 16
  30. 2020Heatmap Regression via Randomized Roundingcited by 15
  31. 2021HIH: Towards More Accurate Face Alignment via Heatmap in Heatmapcited by 15
  32. 2021Subpixel Heatmap Regression for Facial Landmark Localizationcited by 14
  33. 2020Driver Drowsiness Detection Model Using Convolutional Neural Networks Techniques for Android Applicationcited by 4
  34. 2015Deep Recurrent Regression for Facial Landmark Detectioncited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 7; 2024: 11; 2025: 5; 2026: 3.

26 papers.

Sort:

By area:

  1. 2023One-Step Diffusion Distillation via Deep Equilibrium Modelscited by 60
  2. 2024FaceXFormer: A Unified Transformer for Facial Analysiscited by 34
  3. 2024Deep Equilibrium Models are Almost Equivalent to Not-so-deep Explicit Models for High-dimensional Gaussian Mixturescited by 6
  4. 2024Cascaded Dual Vision Transformer for Accurate Facial Landmark Detectioncited by 5
  5. 2023Test like you Train in Implicit Deep Learningcited by 3
  6. 2023Automated Detection of Cat Facial Landmarkscited by 3
  7. 2024Subhomogeneous Deep Equilibrium Modelscited by 3
  8. 2023Self-supervised Adaptive Pre-training of Multilingual Speech Models for Language and Dialect Identificationcited by 2
  9. 2025Recovering Pulse Waves from Video Using Deep Unrolling and Deep Equilibrium Modelscited by 2
  10. 2025Local Scale Equivariance with Latent Deep Equilibrium Canonicalizercited by 2
  11. 2024ORFormer: Occlusion-Robust Transformer for Accurate Facial Landmark Detectioncited by 1
  12. 2025Heatmap Regression without Soft-Argmax for Facial Landmark Detectioncited by 1
14 more
  1. 2025Event-based Facial Keypoint Alignment via Cross-Modal Fusion Attention and Self-Supervised Multi-Event Representation Learningcited by 1
  2. 2026Lipschitz Multiscale Deep Equilibrium Models: A Theoretically Guaranteed and Accelerated Approachcited by 1
  3. 2023Enhancing Multilingual Speech Recognition through Language Prompt Tuning and Frame-Level Language Adaptercited by 0
  4. 2023GMTR: Graph Matching Transformerscited by 0
  5. 2023DarkShot: Lighting Dark Images with Low-Compute and High-Qualitycited by 0
  6. 2024TRLS: A Time Series Representation Learning Framework via Spectrogram for Medical Signal Processingcited by 0
  7. 2024TEN-GUARD: Tensor Decomposition for Backdoor Attack Detection in Deep Neural Networkscited by 0
  8. 2024T-Pixel2Mesh: Combining Global and Local Transformer for 3D Mesh Generation from a Single Imagecited by 0
  9. 2024Strategic Arms with Side Communication Prevail Over Low-Regret MAB Algorithmscited by 0
  10. 2024Towards General Deepfake Detection with Dynamic Curriculumcited by 0
  11. 2024Attr-Int: A Simple and Effective Entity Alignment Framework for Heterogeneous Knowledge Graphscited by 0
  12. 2025Mask6D: Masked Pose Priors For 6D Object Pose Estimationcited by 0
  13. 2026Uncertainty-quantified Pulse Signal Recovery from Facial Video using Regularized Stochastic Interpolantscited by 0
  14. 2026FAHCD-Net: Frequency-Adaptive Heatmap-Conditional Diffusion Networks for Robust Facial Landmark Detectioncited by 0