mlsanityData to 14 Oct 2026

Real-time Monocular Depth Estimation with Sparse Supervision on Mobile

arXiv 2105.12053 ↗First posted 25 May 2021cs.CVcited by 12

Not on the map. Why?

Abstract (arXiv)

Monocular (relative or metric) depth estimation is a critical task for various applications, such as autonomous vehicles, augmented reality and image editing. In recent years, with the increasing availability of mobile devices, accurate and mobile-friendly depth models have gained importance. Increasingly accurate models typically require more computational resources, which inhibits the use of such models on mobile devices. The mobile use case is arguably the most unrestricted one, which requires highly accurate yet mobile-friendly architectures. Therefore, we try to answer the following question: How can we improve a model without adding further complexity (i.e. parameters)? Towards this end, we systematically explore the design space of a relative depth estimation model from various dimensions and we show, with key design choices and ablation studies, even an existing architecture can reach highly competitive performance to the state of the art, with a fraction of the complexity. Our study spans an in-depth backbone model selection process, knowledge distillation, intermediate predictions, model pruning and loss rebalancing. We show that our model, using only DIW as the supervisory dataset, achieves 0.1156 WHDR on DIW with 2.6M parameters and reaches 37 FPS on a mobile GPU, without pruning or hardware-specific optimization. A pruned version of our model achieves 0.1208 WHDR on DIW with 1M parameters and reaches 44 FPS on a mobile GPU.

Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Builds on 36 of its 51 references that are in the map. First posted on arXiv 25 May 2021.

Builds on

36 of 51 references.

References · 36

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  4. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  5. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,442
  6. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  7. 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,611
  8. 2019Searching for MobileNetV3cited by 1,994
  9. 2019Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfercited by 1,495
  10. 2014Depth Map Prediction from a Single Image using a Multi-Scale Deep Networkcited by 1,274
  11. 2018ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Designcited by 1,136
  12. 2018MnasNet: Platform-Aware Neural Architecture Search for Mobilecited by 943
24 more
  1. 2018MegaDepth: Learning Single-View Depth Prediction from Internet Photoscited by 737
  2. 2019Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillationcited by 497
  3. 2018FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Searchcited by 485
  4. 2014Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-Scale Convolutional Architecturecited by 483
  5. 2019GhostNet: More Features from Cheap Operationscited by 410
  6. 2016Deeper Depth Prediction with Fully Convolutional Residual Networkscited by 403
  7. 2016Unsupervised CNN for Single View Depth Estimation: Geometry to the Rescuecited by 368
  8. 2019Enforcing geometric constraints of virtual normal for depth predictioncited by 234
  9. 2020On the uncertainty of self-supervised monocular depth estimationcited by 182
  10. 2016Single-Image Depth Perception in the Wildcited by 180
  11. 2018NetAdapt: Platform-Aware Neural Network Adaptation for Mobile Applicationscited by 147
  12. 2018Revisiting Single Image Depth Estimation: Toward Higher Resolution Maps with Accurate Object Boundariescited by 145
  13. 2019Learning the Depths of Moving People by Watching Frozen Peoplecited by 120
  14. 2020Self-supervised Monocular Trained Depth Estimation using Self-attention and Discrete Disparity Volumecited by 116
  15. 2019MixConv: Mixed Depthwise Convolutional Kernelscited by 108
  16. 2014Deep Convolutional Neural Fields for Depth Estimation from a Single Imagecited by 106
  17. 2019FastDepth: Fast Monocular Depth Estimation on Embedded Systemscited by 94
  18. 2018Structured Attention Guided Convolutional Neural Fields for Monocular Depth Estimationcited by 78
  19. 2019DepthTransfer: Depth Extraction from Video Using Non-parametric Samplingcited by 60
  20. 2018Towards real-time unsupervised monocular depth estimation on CPUcited by 26
  21. 2018Learning Single-Image Depth from Videos using Quality Assessment Networkscited by 25
  22. 2020Real-time single image depth perception in the wild with handheld devicescited by 19
  23. 2020MTAdam: Automatic Balancing of Multiple Training Loss Termscited by 13
  24. 2020A New Distributional Ranking Loss With Uncertainty: Illustrated in Relative Depth Estimationcited by 5

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 2; 2023: 4; 2024: 4; 2025: 1.

12 papers.

  1. 2024SharpDepth: Sharpening Metric Depth Predictions Using Diffusion Distillationcited by 21
  2. 2023MobileVOS: Real-Time Video Object Segmentation Contrastive Learning meets Knowledge Distillationcited by 16
  3. 2024Deep Learning-based Depth Estimation Methods from Monocular Image and Videos: A Comprehensive Surveycited by 15
  4. 2024ScaleDepth: Decomposing Metric Depth Estimation into Scale Prediction and Relative Depth Estimationcited by 15
  5. 2023The RoboDepth Challenge: Methods and Advancements Towards Robust Depth Estimationcited by 13
  6. 2021Lightweight Monocular Depth with a Novel Neural Architecture Search Methodcited by 7
  7. 2024METER: a mobile vision transformer architecture for monocular depth estimationcited by 7
  8. 2023LP-IOANet: Efficient High Resolution Document Shadow Removalcited by 5
  9. 2022Adaptive Mask-based Pyramid Network for Realistic Bokeh Renderingcited by 3
  10. 2023Depth-Relative Self Attention for Monocular Depth Estimationcited by 1
  11. 2022Fast Neural Architecture Search for Lightweight Dense Prediction Networkscited by 0
  12. 2025Rethinking Encoder-Decoder Flow Through Shared Structurescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023MonoProb: Self-Supervised Monocular Depth Estimation with Interpretable Uncertainty
  2. 2026ZipDepth: Bringing Lightweight Zero-Shot Monocular Depth Anywhere, on Any Device
  3. 2026RAD: Retrieval-Augmented Monocular Metric Depth Estimation for Underrepresented Classes
  4. 2023Towards Zero-Shot Scale-Aware Monocular Depth Estimation
  5. 2026Marigold V2: Revisiting Diffusion Transformers for Monocular Depth Estimation
  6. 2024Depth Anything V2
  7. 2026Breaking the Horizontal Prior: From Long-Tailed Orientation Bias to Roll-Robust Monocular Depth Estimation
  8. 2023Monocular Depth Estimation using Diffusion Models