Bosphorus Database for 3D Face Analysis

By A. Savran, Nese Alyüz, H. Dibeklioğlu, Oya Celiktutan, B. Gökberk, Bülent Sankur, Lale Akarun. 2008

908 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Generating 3D faces using Convolutional Mesh Autoencoders (2018, background): “Existing large scale datasets [14, 16, 50, 49, 38] do not contain high resolution extreme facial expressions.”
Multilinear Autoencoder for 3D Face Model Learning (2018, methodology): “A first MAE is learned from static data, using the combined Bosphorus and BU3DFE databases, with registrations provided by Bolkart and Wuhrer, for a total of 5194 meshes.”
Deep Face Recognition: A Survey (2018, neutral): “d CASIA [89] 17,580 725 24.2 OC) 98.32%@ 10 3 [177] pose -B1) [90] 2009 5097 202 25.5 OC) 85.00%@ 10 3 [125] and -B1) [166] 1,212 606 2 (CMC) [201] -B3) [204] 2,388 1,194 2 (CMC) [163] ariation; -B3) [132] 4,652 105 31/44.3/54 OC); (CMC) [81] occlusions -D1) [195] 2,500 100 25 OC); (CMC) [81] -D1) [”
Review of statistical shape spaces for 3D data with comparative analysis for human faces (2012, methodology): “It also likely reﬂects the fact that landmarks in the test database [68], were placed digitally by clicking the scan rather than placed physically by an anthropometrist.”
3D Face Reconstruction by Learning from Synthetic Data (2016, methodology): “We used the reconstructions from the Bosphorus dataset [23], augmented with some in-house data, and aligned using a non-rigid ICP [24] procedure.”
Learning Detailed Face Reconstruction from a Single Image (2016, methodology): “Aid and Aexp are collected from the Bosphorus dataset [37] as in [34], where the identity is modeled using 200 coefficients, and the expression using 84.”
Automatic Analysis of Facial Actions: A Survey (2019, neutral): “AM-FED [117] Induced N/A Various head poses D 10 N N Bosphorous [155] On command 105 subjects 3D multi-pose S 25 N Y BP4D [218] Induced 41 subjects 3D multi-pose D 27 N Y CASME [207] Induced (micro) 35 subjects Near frontal D F Y N CASME II [206] Induced (micro) 26 subjects Near frontal D F Y N Cohn”
Capture, Learning, and Synthesis of 3D Speaking Styles (2019, neutral): “[13, 47, 65]) or dynamic facial expressions (e.”
Facial Expression Analysis under Partial Occlusion (2018, neutral): “8) The Bosphorus database [Savran et al. 2008] is a large multi-expression and multi-pose 3D face dataset with different types of real-life face occlusions.”
Multilinear Wavelets: A Statistical Shape Space for Human Faces (2014, methodology): “We fit all three models to 80 scans (20 subjects, 4 types of occlusions) from the Bosphorus database [23].”
Static and dynamic 3D facial expression recognition: A comprehensive survey (2012, methodology): “Another publicly available dataset consisting of static 3D facial models is the Bosphorus database [76].”
OpenFace 2.0: Facial Behavior Analysis Toolkit (2018, methodology): “Our models are trained on DISFA [48], SEMAINE [51], BP4D [80], UNBC-McMaster [44], Bosphorus [59] and FERA 2011 [66] datasets.”
RGBD Datasets: Past, Present and Future (2016, methodology): “scanner 15 expressions performed by one face - ’07 CASIA 3D Face Database [19] 123 Minolta Vivid 910 4624 images of various expressions, poses and lighting Expression performed ’08 Bosphorus Database [4] 105 Inspeck Mega Capturor II 3D Faces performing expressions at different rotations Expression a”
Multimodal Spontaneous Emotion Corpus for Human Behavior Analysis (2016, background): “Researchers are already beginning to use 3D sensors and models to improve facial feature tracking and expression recognition [29, 34, 21, 25, 20, 33, 32].”
The MUG facial expression database (2010, background): “The MMI facial expression database [7] contains mainly posed but also spontaneous expressions of facial behavior.”
BP4D-Spontaneous: a high-resolution spontaneous 3D dynamic facial expression database (2014, methodology): “The CMU Multi-PIE database [34], 3Ddynamic AU database [35], Bosphorus database [9], KDEF [33], BU 3D facial expression databases [14,15], and ICT-3DRFE database [24] begin to address the need for 3D (or multi-view) data but are limited to posed facial behavior.”
Survey on RGB, 3D, Thermal, and Multimodal Approaches for Facial Expression Recognition: History, Trends, and Affect-related Applications (2016, background): “Bosphorus has low ethnic diversity but it contains a much larger number of expressions, different head poses and deliberate occlusions.”
Automatic Analysis of Facial Affect: A Survey of Registration, Representation, and Recognition (2015, background): “from naturalistic environments with spontaneous affective behaviour [118] as existing datasets [122], [138], [176] contain exaggeratedly posed affective behaviour.”
FaceScape: A Large-Scale High Quality 3D Face Dataset and Detailed Riggable 3D Face Prediction (2020, methodology): “To obtain the accurate 3D face shape, a number of works reconstructed the 3D face using active method including depth sensor or scanner[56, 55, 3, 40, 39, 12, 17], while the other works built sparse multi-view camera system[57, 18].”