The DOI matches its Crossref record; 61 of the 61 references it cites were found in a registry or an index.
Separating Style and Content with Bilinear Models
Abstract
Perceptual systems routinely separate “content” from “style,” classifying familiar words spoken in an unfamiliar accent, identifying a font or handwriting style across letters, or recognizing a familiar face or object seen under unfamiliar viewing conditions. Yet a general and tractable computational model of this ability to untangle the underlying factors of perceptual observations remains elusive (Hofstadter, 1985). Existing factor models (Mardia, Kent, & Bibby, 1979; Hinton & Zemel, 1994; Ghahramani, 1995; Bell & Sejnowski, 1995; Hinton, Dayan, Frey, & Neal, 1995; Dayan, Hinton, Neal, & Zemel, 1995; Hinton & Ghahramani, 1997) are either insufficiently rich to capture the complex interactions of perceptually meaningful factors such as phoneme and speaker accent or letter and font, or do not allow efficient learning algorithms. We present a general framework for learning to solve two-factor tasks using bilinear models, which provide sufficiently expressive representations of factor interactions but can nonetheless be fit to data using efficient algorithms based on the singular value decomposition and expectation-maximization. We report promising results on three different tasks in three different perceptual domains: spoken vowel classification with a benchmark multi-speaker database, extrapolation of fonts to unseen letters, and translation of faces to novel illuminants.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“GeAN also uses the bilinear model [29,30] to distil the knowledge interwoven in these representations by allowing each node and edge to collaboratively modulate each other's contributions through a shared learning space.”
From TVGeAN: Tensor Visibility Graph-Enhanced Attention Network for Versatile Multivariant Time Series Learning Tasks (2024) · cites this paper as [29]
Citations
Cites
Maximum Likelihood from Incomplete Data Via the EM Algorithm
Dempster, A. P., Laird, N. M., & Rubin, D. B. (1977) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1977.tb01600.x · Cited by 50,444
Neural networks for pattern recognition
Bishop, C. (1994) · Choice Reviews Online · doi:10.5860/choice.31-5500 · Cited by 18,537
Press, W. H., Teukolsky, S. A., Vetterling, W. T., & Flannery, B. P. (1994) · Cited by 15,411
Turk, M., & Pentland, A. (1991) · Journal of Cognitive Neuroscience · doi:10.1162/jocn.1991.3.1.71 · Cited by 13,733
(1973) · CERN Document Server (European Organization for Nuclear Research) · Cited by 12,895
Numerical recipes in Pascal: the art of scientific computing
(1990) · Choice Reviews Online · doi:10.5860/choice.27-5158 · Cited by 12,040
An Information-Maximization Approach to Blind Separation and Blind Deconvolution
Bell, A. J., & Sejnowski, T. J. (1995) · Neural Computation · doi:10.1162/neco.1995.7.6.1129 · Cited by 8,986
Shape and motion from image streams under orthography: a factorization method
Tomasi, C., & Takeo Kanade, A. (1992) · International Journal of Computer Vision · doi:10.1007/bf00129684 · Cited by 2,879
Matrix Differential Calculus with Applications in Statistics and Econometrics.
Magnus, J. R., & Neudecker, H. (1988) · Biometrics · doi:10.2307/2531754 · Cited by 2,729
Classical Electrodynamics (2nd edn)
Brereton, M. (1976) · Physics Bulletin · doi:10.1088/0031-9112/27/8/041 · Cited by 2,612
Cited by
Sorted by citations
Items 101–200 of 839
Learning bilinear models for two-factor problems in vision
Freeman, W. T., & Tenenbaum, J. B. (2002) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.1997.609380 · Cited by 109
Self-citationCategory-Based Deep CCA for Fine-Grained Venue Discovery From Multimodal Data
Yu, Y., Tang, S., Aizawa, K., & Aizawa, A. (2018) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2856253 · Cited by 108
“CCA [14], cross-modal factor analysis (CFA) [15], Bilinear model [16] as joint dimensionality reduction techniques were proposed to learn linear functions that map different modalities to the same canonical (semantic) space where the correlation between two modalities is maximized.”
MethodModeling View and Posture Manifolds for Tracking
Lee, C.-S., & Elgammal, A. (2007) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2007.4409030 · Cited by 108
“matrix C and then the view factors are decomposed by fitting an asymmetric bilinear model [ 17 ].”
MethodDisentangling Factors of Variation with Cycle-Consistent Variational Auto-encoders
Jha, A. H., Anand, S., Singh, M., & Veeravasarapu, V. S. R. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01219-9_49 · Cited by 115
“Tenenbaum et al. [16] learn bilinear maps from style and content parameters to images.”
BackgroundProbabilistic models for incomplete multi-dimensional arrays
Chu, W., & Zoubin Ghahramani, A. (2009) · Cited by 106
Tang, Z., Tian, E., Wang, Y., Wang, L., & Yang, T. (2020) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2020.2985159 · Cited by 106
“Bilinear models are effective models to build two key factor variations (i.e., style and content) for images [27].”
BackgroundStyle-content separation by anisotropic part scales
Xu, K., Li, H.-H., Zhang, H., Cohen-Or, D., Xiong, Y.-S., & Cheng, Z.-Q. (2010) · ACM Transactions on Graphics · doi:10.1145/1866158.1866206 · Cited by 154
A Maximum-Likelihood Interpretation for Slow Feature Analysis
Turner, R. E., & Sahani, M. (2007) · Neural Computation · doi:10.1162/neco.2007.19.4.1022 · Cited by 107
“Bilinear learning algorithms have recently been developed in which the effects of the two are segregated (Tenenbaum & Freeman, 2000; Grimes & Rao, 2005).”
BackgroundMethodDeep Learning of Representations: Looking ForwardPreprint
Bengio, Y. (2013) · arXiv (Cornell University) · doi:10.48550/arxiv.1305.0445 · Cited by 149
“…priors on the form of the model can be used to enhance disentangling, such as multiplicative interactions between the factors (Tenenbaum and Freeman, 2000; Desjardins et al., 2012) or orthogonality of the features derivative with respect to the input (Rifai et al., 2011b, 2012a; Sohn et…”
MethodStyle-Aware Mid-level Representation for Discovering Visual Connections in Space and Time
Lee, Y. J., Efros, A. A., & Hebert, M. (2013) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2013.233 · Cited by 97
“Visual style analysis The seminal paper on “style-content separation” [26] uses bilinear models to factor out the style and content components in pre-segmented, pre-aligned visual data (e.”
Identifying Team Style in Soccer Using Formations Learned from Spatiotemporal Tracking Data
Bialkowski, A., Lucey, P., Carr, P., Yue, Y., Sridharan, S., & Matthews, I. A. (2014) · IEEE ... International Conference on Data Mining workshops · doi:10.1109/icdmw.2014.167 · Cited by 100
Discriminative Dictionary Learning With Common Label Alignment for Cross-Modal Retrieval
Deng, C. D., Tang, X., Yan, J., Liu, W., & Gao, X. (2015) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2015.2508146 · Cited by 99
Artistic glyph image synthesis via one-stage few-shot learning
Gao, Y., Guo, Y., Lian, Z., Tang, Y., & Xiao, J. (2019) · ACM Transactions on Graphics · doi:10.1145/3355089.3356574 · Cited by 98
“For instance, Tenenbaum et al. [2000] presented a general bilinear model to solve two-factor tasks, providing sufficiently expressive representations of factor interactions.”
BackgroundZhou, S. K., Aggarwal, G., Chellappa, R., et al. (2007) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2007.25 · Cited by 98
“Early works from the SFS literature assume a constant albedo field which is not valid for many real objects, thus limiting the practical applicability of the SFS algorithms.”
BackgroundResultInfluentialExploiting Unrelated Tasks in Multi-Task Learning
Romera‐Paredes, B., Argyriou, A. A., Bianchi‐Berthouze, N., & Pontil, M. (2012) · Cited by 98
Addressing Complex and Subjective Product-Related Queries with Customer Reviews
McAuley, J., & Yang, A. (2016) · Proceedings of the 25th International Conference on World Wide Web · doi:10.1145/2872427.2883044 · Cited by 190
“Bilinear models [8, 12, 42] can help to address this issue by learning complex mappings between words in one corpus and words in another (or more generally between arbitrary feature spaces).”
BackgroundUnified subspace learning for incomplete and unlabeled multi-view data
Yin, Q., Wu, S., & Wang, L. (2017) · Pattern Recognition · doi:10.1016/j.patcog.2017.01.035 · Cited by 97
Li, X., Lei, L., Sun, Y., Li, M., & Kuang, G. (2020) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2020.2975252 · Cited by 101
“Different from thepreviousbilinearfusionwork,suchasBCNN[30],weusethe SACSMtoexploretheinterchannelrelationshipofthefeature maps X ∈ R h (cid:6) × w (cid:6) × c and Y ∈ R h (cid:6) × w (cid:6) × c , respectively.”
BackgroundMethodInfluentialBayesian Network Learning with Parameter Constraints
Niculescu, R. S., Mitchell, T. M., & Rao, R. B. (2006) · doi:10.5555/1248547.1248597 · Cited by 96
Style-content separation by anisotropic part scales
Xu, K., Li, H., Zhang, H., Cohen‐Or, D., Xiong, Y., & Cheng, Z. (2010) · ACM Transactions on Graphics · doi:10.1145/1882261.1866206 · Cited by 96
A similarity measure for illustration style
Garcés, E., Agarwala, A., Gutiérrez, D., et al. (2014) · ACM Transactions on Graphics · doi:10.1145/2601097.2601131 · Cited by 95
DualVGR: A Dual-Visual Graph Reasoning Unit for Video Question Answering
Wang, J., Bao, B., & Xu, C. (2021) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2021.3097171 · Cited by 100
“Therefore, some approaches [22] related to Bilin-ear Pooling, which has been used to integrate different CNN features for fine-grained image recognition [23], have been proposed.”
Natural grouping of neural responses reveals spatially segregated clusters in prearcuate cortex
Kiani, R., Cueva, C. J., Reppas, J., Peixoto, D., Ryu, S., & Newsome, W. (2015) · Neuron · doi:10.1016/j.neuron.2015.02.014 · Cited by 96
“We used a nonlinear MDS technique (Isomap), (De’ath, 1999; Tenenbaum and Freeman, 2000), but we obtained similar results with other MDS methods.”
Bilinear Sparse Coding for Invariant Vision
Grimes, D. B., & Rao, R. P. N. (2004) · Neural Computation · doi:10.1162/0899766052530893 · Cited by 92
“Previous approaches used unconstrained bilinear models and produced global basis vectors for image representation (Tenenbaum & Freeman, 2000).”
BackgroundMethodInfluentialRecognition and interpretation of parametric gesture
Wilson, A. D., & Bobick, A. (2002) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.1998.710739 · Cited by 93
A High-Resolution Atlas and Statistical Model of the Human Heart From Multislice CT
Hoogendoorn, C., Duchateau, N., Sánchez‐Quintana, D., Whitmarsh, T., Sukno, F. M., Craene, M. D., Lekadir, K., & Frangi, A. F. (2012) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2012.2230015 · Cited by 91
“Training of a bilinear shape model is done as described in [81], [17].”
MethodLun, Z., Kalogerakis, E., & Sheffer, A. (2015) · ACM Transactions on Graphics · doi:10.1145/2766929 · Cited by 88
Memisevic, R. (2013) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2013.53 · Cited by 98
“Like an auto-encoder, this model can be trained with gradient based optimization.”
BackgroundMethodMemisevic, R., Zach, C., Pollefeys, M., & Hinton, G. E. (2010) · Cited by 90
Facial expression space learning
Chuang, E. S., Deshpande, F., & Bregler, C. (2003) · Pacific Conference on Computer Graphics and Applications · doi:10.1109/pccga.2002.1167840 · Cited by 90
“We propose using the factorization techniques that have been studied extensively in the statistical learning, pattern recognition, and computer vision communities [14][15][29][30][33].”
MethodModality-Dependent Cross-Media Retrieval
Wei, Y., Zhao, Y., Zhu, Z., et al. (2016) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/2775109 · Cited by 89
Mohammed, U., Prince, S. J. D., & Kautz, J. (2009) · ACM Transactions on Graphics · doi:10.1145/1531326.1531363 · Cited by 89
Attribute-Enhanced Face Recognition with Neural Tensor Fusion Networks
Hu, G., Yang, H., Yuan, Y., Zhang, Z., Lu, Z., Mukherjee, S. S., Hospedales, T. M., Robertson, N. M., & Yang, Y. (2017) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.404 · Cited by 86
“Another three unsupervised fusion methods: score fusion, CCA [35] and BLM [45] achieve the accuracy between using LeanFace only and AttNet only.”
BackgroundMethodLearning High Dimensional Correspondences from Low Dimensional Manifolds
Ham, J., Lee, D. D., & Saul, L. K. (2003) · Scholarly Commons (University of Pennsylvania) · Cited by 84
Learning Intermediate-Level Representations of Form and Motion from Natural Movies
Cadieu, C. F., & Olshausen, B. A. (2011) · Neural Computation · doi:10.1162/neco_a_00247 · Cited by 87
Simple substrates for complex cognition
Dayan, P. (2008) · Frontiers in Neuroscience · doi:10.3389/neuro.01.031.2008 · Cited by 84
“…the random hidden units of Rigotti and Fusi (2006) and Rigotti et al. (2007), this model used a bilinear form (Koenderink and Van Doorn, 1997; Tenenbaum and Freeman, 2000), which is a rather simpler structure that is an abstraction of the sort of multidimensional basis functions that are…”
MethodBilinear heterogeneous information machine for RGB-D action recognition
Kong, Y., & Fu, Y. (2015) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2015.7298708 · Cited by 82
“Moreover, they use features in a vector form, in which spatiotemporal structures would easily collapse [18, 8].”
MethodRalph Gross, A., Sweeney, L., Cohn, J. F., de la Torre, F. N., & Baker, S. (2009) · Springer eBooks · doi:10.1007/978-1-84882-301-3_8 · Cited by 81
“(Note that for bilinear models and fully labeled data, the iterative Tenenbaum–Freeman algorithm can be used [33].”
MethodA Tensor Approximation Approach to Dimensionality Reduction
Wang, H., & Ahuja, N. (2007) · International Journal of Computer Vision · doi:10.1007/s11263-007-0053-0 · Cited by 83
“Tenenbaum and Freeman (2000) presented a bilinear model for separating the underlying two factors (so-called content and style) of perceptual observations.”
BackgroundInformation fusion in visual question answering: A Survey
Zhang, D., Cao, R., & Wu, S. (2019) · Information Fusion · doi:10.1016/j.inffus.2019.03.005 · Cited by 82
Multi-Modal Machine Learning in Engineering Design: A Review and Future Directions
Song, B., Zhou, R., & Ahmed, F. (2023) · Journal of Computing and Information Science in Engineering · doi:10.1115/1.4063954 · Cited by 105
Writer Adaptation with Style Transfer Mapping
Zhang, X.-Y., & Liu, C. (2012) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2012.239 · Cited by 79
“Tenenbaum and Freeman proposed a bilinear model to separate the style and content (class) information in a group of patterns [35].”
BackgroundSTEFANN: Scene Text Editor Using Font Adaptive Neural Network
Roy, P., Bhattacharya, S., Ghosh, S., & Pal, U. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.01324 · Cited by 84
“Along with statistical models [24] and bilinear factorization [30], machine learning algorithms are used to transfer font features.”
MethodSee the Sound, Hear the Pixels
Ramaswamy, J., & Das, S. (2020) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv45572.2020.9093616 · Cited by 75
“AVFB uses a hybrid of neural network (LSTM) [25, 26, 32, 33, 49] and multi-modal bilinear pooling [7, 10, 22, 41, 42, 50] to fuse the feature representations obtained after extraction using pre-trained CNNs from the two modalities (that is, from audio of single-channel and video which are broken…”
MethodTimbreTron: A WaveNet(CycleGAN(CQT(Audio))) Pipeline for Musical Timbre TransferPreprint
Huang, S., Li, Q., Anil, C., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.09620 · Cited by 77
“In a seminal work, Tenenbaum and Freeman (1999) used a multilinear representation to separate style and content of images.”
BackgroundPei, X., Zuo, K., Li, Y., & Pang, Z. (2023) · International Journal of Computational Intelligence Systems · doi:10.1007/s44196-023-00225-6 · Cited by 77
“The multi-modal fusion method-based tensors, also known as the bilinear pooling fusion method, are predominantly used to fuse visual feature vectors and text feature vectors to attain a joint representation space [33].”
MethodMultilinear Principal Component Analysis of Tensor Objects for Recognition
Lu, H., Plataniotis, K. N., & Venetsanopoulos, A. N. (2006) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2006.837 · Cited by 76
“Bilinear models were proposed to separate style and content in [ 10 ].”
MethodElgammal, A., & Lee, C.-S. (2009) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2008.101 · Cited by 76
“(c) Visualization of a torus manifold with the trajectories of the three different views shown in (a). procedure and can be achieved by adding a regularization term to the diagonal of the matrix ½ ðx1Þ; . . . ; ðxKÞ [60].”
BackgroundAutomatic 3D Facial Expression Analysis in Videos
Chang, Y., Vieira, M., Turk, M., & Velho, L. (2005) · Lecture Notes in Computer Science · doi:10.1007/11564386_23 · Cited by 74
“Considering these factors, bilinear [7] and multi-linear [8] models have been successful in decomposing the static image ensembles into different sources of variation, such as identity and content.”
MethodLee, S., Cho, S., & Im, S. (2021) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr46437.2021.01500 · Cited by 80
“The separation of style and content components in a latent space has been widely studied for artistic style transfer [33, 8, 11, 42, 43].”
BackgroundRobust pose invariant face recognition using coupled latent space discriminant analysis
Sharma, A., Haj, M. A., Choi, J., et al. (2012) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2012.08.001 · Cited by 74
Style consistent classification of isogenous patterns
Sarkar, P., & George Nagy, A. (2004) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2005.18 · Cited by 74
Large-Scale Category Structure Aware Image Categorization
Zhao, B., Fei-Fei, L., & Xing, E. P. (2011) · Cited by 74
Learning Visual Context for Group Activity Recognition
Yuan, H., & Ni, D. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i4.16437 · Cited by 73
“Bilinear Pooling Bilinear pooling was firstly proposed to provide a sufficiently expressive representations of two-factor interactions (Tenenbaum and Freeman 2000).”
BackgroundCross-Modal Subspace Learning via Pairwise Constraints
He, R., Zhang, M., Wang, L., et al. (2015) · IEEE Transactions on Image Processing · doi:10.1109/tip.2015.2466106 · Cited by 73
An End-to-End Multimodal Voice Activity Detection Using WaveNet Encoder and Residual Networks
Ariav, I., & Cohen, I. (2019) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2019.2901195 · Cited by 74
Manifold Based Analysis of Facial Expression
Hu, C., Chang, Y. H., Feris, R. S., & Turk, M. (2005) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvpr.2004.390 · Cited by 74
A Multi-Task Approach for Disentangling Syntax and Semantics in Sentence Representations
Chen, M., Tang, Q., Wiseman, S., & Gimpel, K. (2019) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/n19-1254 · Cited by 75
“There is a growing amount of work on learning interpretable or disentangled latent representations both in machine learning (Tenenbaum and Freeman, 2000; Reed et al., 2014; Makhzani et al., 2015; Mathieu et al., 2016; Higgins et al., 2016; Chen et al., 2016; Hsu et al., 2017) and in various NLP applications, including…”
BackgroundFDDH: Fast Discriminative Discrete Hashing for Large-Scale Cross-Modal Retrieval
Liu, X., Wang, X., & Cheung, Y. (2021) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2021.3076684 · Cited by 71
“Similarly, partial least-squares (PLS) [18] and bilinear model (BLM) [19] also learn a common latent sub-”
BackgroundTwitter100k: A Real-World Dataset for Weakly Supervised Cross-Media Retrieval
Hu, Y., Zheng, L., Yang, Y., & Huang, Y. (2017) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2017.2760101 · Cited by 71
“As a minor contribution, inspired by the characteristic of Twitter100k, we propose an OCR-based cross-media retrieval method.”
MethodComputer Facial Animation: A Survey
Deng, Z., & Noh, J. (2007) · Springer eBooks · doi:10.1007/978-1-84628-907-1_1 · Cited by 74
“Chuang and Bregler [116, 117] learn a facial expression mapping/transformation function from training video footage using bilinear models [118], and then this learned mapping is used to transform input video of neutral talking to expressive talking.”
MethodBilinear Modeling via Augmented Lagrange Multipliers (BALM)
Bue, A. D., Xavier, J., Agapito, L., et al. (2011) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2011.238 · Cited by 69
“The interaction between two factors has also been generalized to several problems [6] in the machine learning field, perhaps the closest computational interpretation of Schlick's phylosophical analysis [7] asserting that content can be only formed and expressed through the subjectivity of the…”
BackgroundEmploying Bilinear Fusion and Saliency Prior Information for RGB-D Salient Object Detection
Huang, N., Yang, Y., Zhang, D., Zhang, Q., & Han, J. (2021) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2021.3069297 · Cited by 71
“Bilinear Model [38], [39] has been widely studied in traditional machine learning due to its powerful representation ability.”
BackgroundMethodStyle translation for human motion
Hsu, E., Pulli, K., & Popović, J. (2005) · ACM SIGGRAPH Conference (SIGGRAPH) · doi:10.1145/1186822.1073315 · Cited by 72
“To show the quality of correspondences independently of translation, we extracted single gait cycles from the aligned motions, transformed them to the representation described in §4.1, and applied bilinear analysis to separate style and speed [Vasilescu 2002; Tenenbaum and Freeman 2000].”
MethodInfluentialFA-GAN: Face Augmentation GAN for Deformation-Invariant Face Recognition
Luo, M., Cao, J., Ma, X., Zhang, X., & He, R. (2021) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2021.3053460 · Cited by 68
“Disentangled representation learning [17], [66]–[71] aims to model the key factors that affect the shape of the data so that changes in a key factor only cause changes in only a certain feature of the data, while other features are not affected.”
BackgroundBilinearity, rules, and prefrontal cortex
Dayan, P. (2007) · Frontiers in Computational Neuroscience · doi:10.3389/neuro.10.001.2007 · Cited by 67
“However, as also implied in the work of Dayan (2006) and Tenenbaum and Freeman (2000), we suggest that this abstraction is useful, since it allows us to focus on the relationship between rules and habits, and, in the longer run, to consider the status of rules with respect to the overall…”
BackgroundMethodDisentangling Factors of Variation via Generative EntanglingPreprint
Desjardins, G., Courville, A. C., & Bengio, Y. (2012) · arXiv (Cornell University) · doi:10.48550/arxiv.1210.5474 · Cited by 66
“Formally, the elements of observation x are given by x k = i j W ijk y i z j , ∀ k , where y i and z j are elements of the two factors ( y and z ) representing the observation and W ijk is an element of the tensor of model parameters (Tenenbaum and Freeman, 2000).”
BackgroundMethodInfluentialCoupled Visual and Kinematic Manifold Models for Tracking
Lee, C.-S., & Elgammal, A. (2009) · International Journal of Computer Vision · doi:10.1007/s11263-009-0266-5 · Cited by 65
“After selecting the best matching, we further estimated the view and style parameters.”
MethodInfluentialSemi-supervised learning of multi-factor models for face de-identification
Ralph Gross, A., Sweeney, L., la Torre, F. D., & Baker, S. (2008) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2008.4587369 · Cited by 68
“Previously proposed algorithms to fit multi-factor models such as the Tenenbaum-Freeman algorithm [20] require complete labels for all training examples.”
MethodInfluentialDeep Learning of Representations
Bengio, Y., & Courville, A. C. (2013) · Intelligent systems reference library · doi:10.1007/978-3-642-36657-4_1 · Cited by 66
““Separating Style and Content with Bilinear Models” Tenenbaum and Freeman (2000).”
BackgroundWhat Is the Other 85 Percent of V1 Doing?
Olshausen, B. A., & Field, D. J. (2006) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780195148220.003.0010 · Cited by 66
GlyphGAN: Style-consistent font generation based on generative adversarial networks
Hayashi, H., Abe, K., & Uchida, S. (2019) · Knowledge-Based Systems · doi:10.1016/j.knosys.2019.104927 · Cited by 64
“One of the classical methods is example-based font generation [8, 9, 10, 11, 12, 13].”
MethodContext-Aware Code Change Embedding for Better Patch Correctness Assessment
Lin, B., Wang, S., Wen, M., & Mao, X. (2022) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3505247 · Cited by 63
“⑤ E bi : Bilinear model , which is proposed by Tanenbaum et al. [75] and has been shown to be able to efficiently combine the features of two vectors.”
BackgroundResultKumar, G. K., & Nandakumar, K. (2022) · Workshop on NLP for Positive Impact (NLP4PI) · doi:10.18653/v1/2022.nlp4pi-1.20 · Cited by 111
Bilinear CNNs for Fine-grained Visual RecognitionPreprint
Lin, T., RoyChowdhury, A., & Maji, S. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1504.07889 · Cited by 62
“Bilinear models were proposed by Tanenbaum and Freeman [33] to model two-factor variations, such as “style” and “content”, for images.”
MethodProbabilistic Video Generation Using Holistic Attribute Control
He, J., Lehrmann, A., Marino, J., Mori, G., & Sigal, L. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01228-1_28 · Cited by 65
“Our approach is implicitly related to the rich literature on stylecontent separation (a problem introduced in [17]); in our case a distribution over the content, in each frame, is being parameterized by attribute factors that affect the latent state and style is modelled by the motion patterns that result from dynamic…”
BackgroundLearning style translation for the lines of a drawing
Freeman, W. T., Tenenbaum, J. B., & Pasztor, E. C. (2003) · ACM Transactions on Graphics · doi:10.1145/588272.588277 · Cited by 61
“the bilinear models used in Tenenbaum and Freeman [2000] to separate style and content for various synthesis and analysis problems.”
BackgroundMethodSelf-citationRGB-D-Based Object Recognition Using Multimodal Convolutional Neural Networks: A Survey
Gao, M., Jiang, J., Zou, G., John, V., & Liu, Z. (2019) · IEEE Access · doi:10.1109/access.2019.2907071 · Cited by 63
“The activations of the feature extractors for both streamswere combined in a cross modalities fashion via a bilinear operation [157] at various feature scales.”
MethodLearning content and style: Joint action recognition and person identification from human skeletons
Wang, H., & Wang, L. (2018) · Pattern Recognition · doi:10.1016/j.patcog.2018.03.030 · Cited by 56
A case study of behavior-driven conjoint analysis on Yahoo!: front page today module
Chu, W., Park, S.-T., Beaupre, T., Motgi, N., Phadke, A., Chakraborty, S., & Zachariah, J. (2009) · Knowledge Discovery and Data Mining · doi:10.1145/1557019.1557138 · Cited by 84
“For example, Tenenbaum and Freeman [18] developed a bilinear model for separating “style” and “content” in images, and recently Chu and Ghahramani [3] derived a probabilistic framework of the Tucker family for modeling structural dependency from partially observed highdimensional array data.”
BackgroundGradient-based learning of higher-order image features
Memisevic, R. (2011) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2011.6126419 · Cited by 60
“When using a single hidden layer with linear activations, the model takes the form of a (conditionally trained) bi-linear model [31] [8], [20], where additional “simple cells” can be used to pre-project inputs before computing multiplicative interactions.”
BackgroundMethodAffinity regression predicts the recognition code of nucleic acid–binding proteins
Pelossof, R. A., Singh, I., Yang, J. L., Weirauch, M. T., Hughes, T. R., & Leslie, C. S. (2015) · Nature Biotechnology · doi:10.1038/nbt.3343 · Cited by 57
“We can interpret the affinity regression model through mappings to its feature spaces 15 .”
TransMoMo: Invariance-Driven Unsupervised Video Motion Retargeting
Yang, Z., Zhu, W., Wu, W., Qian, C., Zhou, Q., Zhou, B., & Loy, C. C. (2020) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr42600.2020.00535 · Cited by 57
Curvature Generation in Curved Spaces for Few-Shot Learning
Gao, Z. F., Wu, Y., Jia, Y., & Harandi, M. T. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00857 · Cited by 56
“…( (cid:37) is a hyperparameter of the class-curvature generator), and its l -th element b lj can be computed via the factorized bilinear model [24, 46]: where 1 ∈ R r is a vector whose elements are all ‘ 1 ’s. r is the other hyperparameter of the class-curvature generator de-noting the rank of U…”
BackgroundLearning Stylistic Dynamic Movement Primitives from multiple demonstrations
Matsubara, T., Hyon, S.-H., & Morimoto, J. (2010) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2010.5651049 · Cited by 56
“Since such approaches basically lead to a two-factor model that consists of independent control variables for style and content, it can be a useful tool for many applications of the character animation synthesis.”
MethodLin, J., Chen, Z., Xia, Y., et al. (2019) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2019.2950198 · Cited by 55
Cross-Modality Submodular Dictionary Learning for Information Retrieval
Zhu, F., Shao, L., & Yu, M. (2014) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/2661829.2661926 · Cited by 54
“…transfer method (PCA), correlation matching (CM) [ 31 ] , semantic matching (SM) [ 31 ] , semantic correlation matching (SCM) [ 7 ] , a bilinear model (BLM) [ 42 ] , generalized multiview linear discriminant analysis (GMLDA) [ 36 ] andgeneralizedmultiviewmarginalfisheranalysis(GMMFA) [ 36 ] .”
Nonlinear manifold learning for dynamic shape and dynamic appearance
Elgammal, A., & Lee, C.-S. (2007) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2005.09.010 · Cited by 53
“Such subspace analysis can be further extended to decompose other factors using bilinear models and multi-linear tensor analysis [11,12].”
BackgroundJodoin, P., Epstein, E., Granger-Piché, M., et al. (2002) · International Symposium on Non-Photorealistic Animation and Rendering (NPAR) · doi:10.1145/508530.508536 · Cited by 53
Learning the Lie Groups of Visual Invariance
Xu, M., & Rao, R. P. N. (2007) · Neural Computation · doi:10.1162/neco.2007.19.10.2665 · Cited by 53
“A number of other approaches have since been explored (Fukushima, 1980; Hinton, 1987; LeCun et al., 1989; Olshausen, Anderson, & Essen, 1995; Frey & Jojic, 1999; Tenenbaum & Freeman, 2000; Vasilescu & Terzopoulos, 2002; Grimes & Rao, 2005), some relying on pooling of activities in a featuredetector hierarchy (e.g., Fu…”
BackgroundCompensating for speaker or lexical variabilities in speech for emotion recognition
Mariooryad, S., & Busso, C. (2013) · Speech Communication · doi:10.1016/j.specom.2013.07.011 · Cited by 52
“Alternatively, we are planning to study factor analysis models such as bilinear model (Tenenbaum and Freeman, 2000) and tied factor analysis based on probabilistic linear discriminant analysis (PLDA) (Prince et al.”
MethodLearning to Traverse Image Manifolds
Dollár, P., Rabaud, V., & Belongie, S. (2007) · The MIT Press eBooks · doi:10.7551/mitpress/7503.003.0050 · Cited by 52
“The applications range from video texture synthesis [18] and facial expression extrapolation [8, 23] to face recognition [10] and video rewrite [7].”
BackgroundGeneralized Coupled Dictionary Learning Approach With Applications to Cross-Modal Matching
Mandal, D., & Biswas, S. (2016) · IEEE Transactions on Image Processing · doi:10.1109/tip.2016.2577885 · Cited by 52
“Tang & Wang [36] PLS [15] Bilinear [37] SCDL [17] Yang.”
BackgroundLearning unified binary codes for cross-modal retrieval via latent semantic hashing
Xu, X., He, L., Shimada, A., et al. (2016) · Neurocomputing · doi:10.1016/j.neucom.2015.11.133 · Cited by 51
Identity–Expression Dual Branch Network for Facial Expression Recognition
Zhang, H., Su, W., Yu, J., & Wang, Z. (2020) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2020.3034807 · Cited by 52
Bilinear Kernel Reduced Rank Regression for Facial Expression Synthesis
Huang, D., & la Torre, F. D. (2010) · Lecture notes in computer science · doi:10.1007/978-3-642-15552-9_27 · Cited by 50
“Tensor-based approaches [10,11,12] perform Higher-Order Singular Value Decomposition (HOSVD) to factorize the normalized face appearance into identity, expression, pose, and illumination.”
MethodCritic-driven ensemble classification
Miller, D. J., & Yan, L. (1999) · IEEE Transactions on Signal Processing · doi:10.1109/78.790663 · Cited by 50
Enhancing image and video retrieval: learning via equivalence constraints
Hertz, T., Shental, N., Bar-Hillel, A., et al. (2003) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2003.1211531 · Cited by 50
“Note, however, that negative equivalence constraints violate this assumption, as dependencies between the hidden variables are introduced.”
BackgroundGaussian Process Models for Link Analysis and Transfer Learning
Yu, K., & Chu, W. (2007) · neural information processing systems · Cited by 50
Meng, Q., Matthew, J., Zimmer, V. A., Gómez, A., Lloyd, D. F. A., Rueckert, D., & Kainz, B. (2020) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2020.3035424 · Cited by 48
“Traditional models include techniques such as Independent Component Analysis (ICA) [26] and bilinear models [60] as well as learningbased models such as InfoGAN [13] and β-VAE [9, 23].”
Method
Metadata
- Authors
- Tenenbaum, J. B., & Freeman, W. T.
- Affiliations
- Massachusetts Institute of Technology (US)
- Mitsubishi Electric Research Laboratories (United States) (US)
- Venue
- Neural Computation
- Publisher
- MIT Press
- Published
- 2000-06-01
- Volume
- 12
- Issue
- 6
- Pages
- 1247-1283
- ISSN
- 0899-7667 (ISSN-L)
- 1530-888X
- OpenAlex
- W2170653751
- Semantic Scholar
- 7e85f7d59e37972ec52cbabfef0512588d87f125
- Cited by
- 840
- References
- 61
- Citations by year
- 2026: 2 · 2025: 17 · 2024: 20 · 2023: 31 · 2022: 38 · 2021: 61 · 2020: 58 · 2019: 70 · 2018: 49 · 2017: 39 · 2016: 56 · 2015: 34 · 2014: 27 · 2013: 39 · 2012: 30
- FWCI
- 13.29
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Neural Computation (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face and Expression Recognition
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Keywords
- bilinear model, factor models, perceptual learning, speaker accent, singular value decomposition, expectation-maximization algorithm, factor interactions
- MeSH terms
- Algorithms, Face, Humans, Learning, Pattern Recognition, Visual, Speech Perception, Linear Models, Neural Networks, Computer
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
