Factors of Influence for Transfer Learning Across Diverse Appearance Domains and Task Types
Open accessThe DOI matches its Crossref record; 50 of the 50 references it cites were found in a registry or an index.
Abstract
Transfer learning enables to re-use knowledge learned on a source task to help learning a target task. A simple form of transfer learning is common in current state-of-the-art computer vision models, i.e., pre-training a model for image classification on the ILSVRC dataset, and then fine-tune on any target task. However, previous systematic studies of transfer learning have been limited and the circumstances in which it is expected to work are not fully understood. In this paper we carry out an extensive experimental exploration of transfer learning across vastly different image domains (consumer photos, autonomous driving, aerial imagery, underwater, indoor scenes, synthetic, close-ups) and task types (semantic segmentation, object detection, depth estimation, keypoint detection). Importantly, these are all complex, structured output tasks types relevant to modern computer vision applications. In total we carry out over 2000 transfer learning experiments, including many where the source and target come from different image domains, task types, or both. We systematically analyze these experiments to understand the impact of image domain, task type, and dataset size on transfer learning performance. Our study leads to several insights and concrete recommendations: (1) for most tasks there exists a source which significantly outperforms ILSVRC'12 pre-training; (2) the image domain is the most important factor for achieving positive transfer; (3) the source dataset should include the image domain of the target dataset to achieve best results; (4) at the same time, we observe only small negative effects when the image domain of the source task is much broader than that of the target; (5) transfer across task types can be beneficial, but its success is heavily dependent on both the source and target task types.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al. (2020) · arXiv · doi:10.48550/arxiv.2010.11929 · Cited by 47,472
Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
Finn, C., Abbeel, P., & Levine, S. (2017) · arXiv · doi:10.48550/arxiv.1703.03400 · Cited by 8,512 · cited 2×
“Optimization-based methods employ metalearning [28], [70], [74], [75], [107].”
BackgroundOvercoming catastrophic forgetting in neural networks
Kirkpatrick, J., Pascanu, R., Rabinowitz, N., Veness, J., Desjardins, G., Rusu, A. A., Milan, K., Quan, J., Ramalho, T., Grabska-Barwinska, A., et al. (2017) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1611835114 · Cited by 7,899
Zagoruyko, S., & Komodakis, N. (2016) · Procedings of the British Machine Vision Conference 2016 · doi:10.5244/c.30.87 · Cited by 7,244
The Pascal Visual Object Classes Challenge: A Retrospective
Everingham, M., Eslami, S. M. A., Van Gool, L., Williams, C. K. I., Winn, J., & Zisserman, A. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-014-0733-5 · Cited by 7,139
Learning to Compare: Relation Network for Few-Shot Learning
Sung, F., Yang, Y., Zhang, L., Xiang, T., Torr, P. H. S., & Hospedales, T. M. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00131 · Cited by 5,816
Indoor segmentation and support inference from rgbd images
Silberman, N., Hoiem, D., Kohli, P., & Fergus, R. (2012) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-33715-4_54 · Cited by 5,760
Stacked Hourglass Networks for Human Pose Estimation
Newell, A., Yang, K., & Deng, J. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46484-8_29 · Cited by 5,634 · cited 1×
“First, it outperforms ResNet [39], ResNeXt [111], Wide ResNet [115], and Stacked Hourglass Networks [68] on three of the tasks we explore: semantic segmentation, keypoint detection, and object detection.”
MethodAdversarial Discriminative Domain Adaptation
Tzeng, E., Hoffman, J., Saenko, K., & Darrell, T. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.316 · Cited by 5,154 · cited 1×
“Some of the most common techniques are minimizing the discrepancy in feature distributions [32], [61], [82], [91], [97], [100], [119], embedding domain alignment layers into a network [13], [63], [79], directly transform images from the target to the source domain [9], [81], [83].”
MethodDeep High-Resolution Representation Learning for Visual Recognition
Wang, J., Sun, K., Cheng, T., Jiang, B., Deng, C., Zhao, Y., Liu, D., Mu, Y., Tan, M., Wang, X., et al. (2021) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2020.2983686 · Cited by 4,894
Learning Transferable Features with Deep Adaptation Networks
Long, M., Cao, Y., Wang, J., & Jordan, M. I. (2015) · arXiv · doi:10.48550/arxiv.1502.02791 · Cited by 4,505
CenterNet: Keypoint Triplets for Object Detection
Duan, K., Bai, S., Xie, L., Qi, H., Huang, Q., & Tian, Q. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00667 · Cited by 4,079
CornerNet: Detecting Objects as Paired Keypoints
Law, H., & Deng, J. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01264-9_45 · Cited by 3,852
Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning
Grill, J.-B., Strub, F., Altché, F., Tallec, C., Richemond, P. H., Buchatskaya, E., Doersch, C., Pires, B. A., Guo, Z. D., Azar, M. G., et al. (2020) · arXiv · doi:10.48550/arxiv.2006.07733 · Cited by 3,846
Deep CORAL: Correlation Alignment for Deep Domain Adaptation
Sun, B., & Saenko, K. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-49409-8_35 · Cited by 3,757 · cited 1×
“Some of the most common techniques are minimizing the discrepancy in feature distributions [32], [61], [82], [91], [97], [100], [119], embedding domain alignment layers into a network [13], [63], [79], directly transform images from the target to the source domain [9], [81], [83].”
MethodExploring Simple Siamese Representation Learning
Chen, X., & He, K. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.01549 · Cited by 3,725
Kuznetsova, A., Rom, H., Alldrin, N., Uijlings, J., Krasin, I., Pont-Tuset, J., Kamali, S., Popov, S., Malloci, M., Kolesnikov, A., et al. (2018) · arXiv · doi:10.48550/arxiv.1811.00982 · Cited by 3,653
Self-citationDepth Map Prediction from a Single Image using a Multi-Scale Deep Network
Eigen, D., Puhrsch, C., & Fergus, R. (2014) · arXiv · doi:10.48550/arxiv.1406.2283 · Cited by 3,442
DOTA: A Large-Scale Dataset for Object Detection in Aerial Images
Xia, G.-S., Bai, X., Ding, J., Zhu, Z., Belongie, S., Luo, J., Datcu, M., Pelillo, M., & Zhang, L. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00418 · Cited by 3,414
Adapting visual category models to new domains
Saenko, K., Kulis, B., Fritz, M., & Darrell, T. (2010) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-15561-1_16 · Cited by 2,847 · cited 2×
“Some of the most common techniques are minimizing the discrepancy in feature distributions [32], [61], [82], [91], [97], [100], [119], embedding domain alignment layers into a network [13], [63], [79], directly transform images from the target to the source domain [9], [81], [83].”
BackgroundMethodBDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning
Yu, F., Chen, H., Wang, X., Xian, W., Chen, Y., Liu, F., Madhavan, V., & Darrell, T. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00271 · Cited by 2,641 · cited 8×
“For example, CityScapes [18] also provides disparity and Berkeley Deep Drive [113] also includes road segmentation annotations.”
BackgroundInfluentialRevisiting Unreasonable Effectiveness of Data in Deep Learning Era
Sun, C., Shrivastava, A., Singh, S., & Gupta, A. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.97 · Cited by 2,484
Deep Hashing Network for Unsupervised Domain Adaptation
Venkateswara, H., Eusebio, J., Chakraborty, S., & Panchanathan, S. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.572 · Cited by 2,171
Deep Ordinal Regression Network for Monocular Depth Estimation
Fu, H., Gong, M., Wang, C., Batmanghelich, K., & Tao, D. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00214 · Cited by 1,991
SUN RGB-D: A RGB-D scene understanding benchmark suite
Song, S., Lichtenberg, S. P., & Xiao, J. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298655 · Cited by 1,943 · cited 9×
“This is explainable since the multi-source model includes both COCO (consumer) and ScanNet (indoor) which are both good sources for SUN RGB-D.”
BackgroundInfluentialKirillov, A., He, K., Girshick, R., Rother, C., & Dollár, P. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00963 · Cited by 1,659
Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks
Bousmalis, K., Silberman, N., Dohan, D., Erhan, D., & Krishnan, D. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.18 · Cited by 1,603 · cited 1×
“Some of the most common techniques are minimizing the discrepancy in feature distributions [32], [61], [82], [91], [97], [100], [119], embedding domain alignment layers into a network [13], [63], [79], directly transform images from the target to the source domain [9], [81], [83].”
MethodThe Role of Context for Object Detection and Semantic Segmentation in the Wild
Mottaghi, R., Chen, X., Liu, X., Cho, N.-G., Lee, S.-W., Fidler, S., Urtasun, R., & Yuille, A. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.119 · Cited by 1,438 · cited 3×
“TABLE 1 Overview of the 20 Datasets Used in This Paper Nr Name Reference Domain Semantic Detection Keypoints Depth # Train # Val 1 Pascal Context [65] Consumer 60 - - - 5K 5K 2 Pascal VOC [27] Consumer 22 20 - - 10K 1449 a 3 MPII Human Pose [6] Consumer - - 16 - 17K 7K 4 ADE20K [120] Consumer 150 - - - 20K 2K 5 COCO P…”
The Mapillary Vistas Dataset for Semantic Understanding of Street Scenes
Neuhold, G., Ollmann, T., Bulo, S. R., & Kontschieder, P. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.534 · Cited by 1,434 · cited 2×
“TABLE 1 Overview of the 20 Datasets Used in This Paper Nr Name Reference Domain Semantic Detection Keypoints Depth # Train # Val 1 Pascal Context [65] Consumer 60 - - - 5K 5K 2 Pascal VOC [27] Consumer 22 20 - - 10K 1449 a 3 MPII Human Pose [6] Consumer - - 16 - 17K 7K 4 ADE20K [120] Consumer 150 - - - 20K 2K 5 COCO P…”
Do Better ImageNet Models Transfer Better?
Kornblith, S., Shlens, J., & Le, Q. V. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00277 · Cited by 1,325
Dynamic Few-Shot Visual Learning Without Forgetting
Gidaris, S., & Komodakis, N. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00459 · Cited by 1,128
Taskonomy: Disentangling Task Transfer Learning
Zamir, A. R., Sax, A., Shen, W., Guibas, L., Malik, J., & Savarese, S. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00391 · Cited by 1,128
AutoAugment: Learning Augmentation Policies from Data
Cubuk, E. D., Zoph, B., Mane, D., Vasudevan, V., & Le, Q. V. (2018) · arXiv · doi:10.48550/arxiv.1805.09501 · Cited by 1,106
VirtualWorlds as Proxy for Multi-object Tracking Analysis
Gaidon, A., Wang, Q., Cabon, Y., & Vig, E. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.470 · Cited by 1,079
Big Transfer (BiT): General Visual Representation Learning
Kolesnikov, A., Beyer, L., Zhai, X., Puigcerver, J., Yung, J., Gelly, S., & Houlsby, N. (2019) · arXiv · doi:10.48550/arxiv.1912.11370 · Cited by 1,039
Big Transfer (BiT): General Visual Representation Learning
Kolesnikov, A., Beyer, L., Zhai, X., Puigcerver, J., Yung, J., Gelly, S., & Houlsby, N. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58558-7_29 · Cited by 1,039 · cited 8×
“Several works [23], [52], [66], [117] report experiments on target datasets spanning different image domains, especially since the advent of the VTAB suite [117] which assembles datasets captured using a standard camera, as well as”
BackgroundMethodResultInfluentialRethinking Few-Shot Image Classification: A Good Embedding is All You Need?
Tian, Y., Wang, Y., Krishnan, D., Tenenbaum, J. B., & Isola, P. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58568-6_16 · Cited by 952
Kokkinos, I. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.579 · Cited by 657
Learning multiple visual domains with residual adapters
Rebuffi, S.-A., Bilen, H., & Vedaldi, A. (2017) · arXiv · doi:10.48550/arxiv.1705.08045 · Cited by 605 · cited 1×
“for robust vision approaches [3], [58], [76].”
Incremental Learning of Object Detectors without Catastrophic Forgetting
Shmelkov, K., Schmid, C., & Alahari, K. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.368 · Cited by 559
Low-Shot Learning with Imprinted Weights
Qi, H., Brown, M., & Lowe, D. G. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00610 · Cited by 539 · cited 1×
“Overall, the community seems to be reaching a consensus [15], [34], [41], [73], [90], [95]: the”
BackgroundDepth From Videos in the Wild: Unsupervised Monocular Depth Learning From Unknown Cameras
Gordon, A., Li, H., Jonschkowski, R., & Angelova, A. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00907 · Cited by 389
Exploring Simple Siamese Representation Learning
Chen, X., & He, K. (2020) · arXiv · doi:10.48550/arxiv.2011.10566 · Cited by 381
Factors of Transferability for a Generic ConvNet Representation
Azizpour, H., Razavian, A. S., Sullivan, J., Maki, A., & Carlsson, S. (2016) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2015.2500224 · Cited by 301 · cited 2×
“initializing the backbones of our networks with the weights obtained from image classification pre-training on ILSVRC’12 [7], [17], [35], [38], [42], [53], [85], [121].”
BackgroundMethodZhou, X., Wang, D., & Krähenbühl, P. (2019) · arXiv · doi:10.48550/arxiv.1904.07850 · Cited by 288 · cited 9×
“This strategywas shown to be effective on a wide variety of datasets and task types, including image classification [7], [17], [37], [42], [53], object detection [35], semantic segmentation [85], human pose estimation [38], [118], [121], and depth estimation [16], [29].”
BackgroundMethodResultInfluentialVisDA: A Synthetic-to-Real Benchmark for Visual Domain Adaptation
Peng, X., Usman, B., Kaushik, N., Wang, D., Hoffman, J., & Saenko, K. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2018.00271 · Cited by 196 · cited 3×
“By virtue of dataset collection suites such as VTAB [117] andVisDA [71], recently severalworks in transfer learning and domain adaptation experiment with target datasets spanning a variety of image domains [23], [52], [55], [66], [71], [117], [119].”
BackgroundWho Left the Dogs Out? 3D Animal Reconstruction with Expectation Maximization in the Loop
Biggs, B., Boyne, O., Charles, J., Fitzgibbon, A., & Cipolla, R. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58621-8_12 · Cited by 107
Learning Object-Specific Distance From a Monocular Image
Zhu, J., & Fang, Y. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00394 · Cited by 85 · cited 1×
“The soft-plus activation: log ðexpðxÞ þ 1Þ is a differential clipping to convert the logit values to depth ensuring that the predictions are positive, it is also used in [36], [122].”
Method
Cited by
A Review of Deep Transfer Learning and Recent Advancements
Iman, M., Arabnia, H. R., & Rasheed, K. (2023) · Technologies · doi:10.3390/technologies11020040 · Cited by 680
MultiMAE: Multi-modal Multi-task Masked Autoencoders
Bachmann, R., Mizrahi, D., Atanov, A., & Zamir, A. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19836-6_20 · Cited by 268
In silico proof of principle of machine learning-based antibody design at unconstrained scale
Akbar, R., Robert, P. A., Weber, C. R., Widrich, M., Frank, R., Pavlović, M., Scheffer, L., Chernigovskaya, M., Snapkov, I., Slabodkin, A., et al. (2022) · mAbs · doi:10.1080/19420862.2022.2031482 · Cited by 94
How Well Do Sparse ImageNet Models Transfer?
Iofinova, E., Peste, A., Kurtz, M., & Alistarh, D. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.01195 · Cited by 22
Evaluating the structure of cognitive tasks with transfer learning
Aristimunha, B., de Camargo, R. Y., Pinaya, W. H. L., Chevallier, S., Gramfort, A., & Rommel, C. (2023) · arXiv · doi:10.48550/arxiv.2308.02408 · Cited by 4
“These factors seem to be observed in other areas, such as computer vision [96].”
Background
Metadata
- Authors
- Mensink, T., Uijlings, J., Kuznetsova, A., Gygli, M., & Ferrari, V.
- Venue
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Published
- 2022-12-01
- Volume
- 44
- Issue
- 12
- Pages
- 9298-9314
- ISSN
- 0162-8828 (ISSN-L)
- 2160-9292
- 1939-3539
- OpenAlex
- W3135975972
- Cited by
- 80
- References
- 50
- FWCI
- 5.03
- Type
- journal article
- Language
- en
- Open access
- hybrid
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.