Columbia Object Image Library (COIL100)

By Sameer A. Nene, Shree K. Nayar, Hiroshi Murase. 1996

Subfields: Computer Vision and Pattern Recognition, Evaluation & Benchmarks

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

In Evaluation & Benchmarks

Citations in the corpus

UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction (2018, methodology): “For example, Figure 1 shows both UMAP and t-SNE embeddings of the COIL20, MNIST, Fashion MNIST, and Google News datasets.”
Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shift (2018, background): “Speciﬁcally, we explore the following types of shifts: × × [34] where images between 0 ◦ and 175 ◦ are sampled by the source and images between 180 ◦ and 355 ◦ are sampled by the target distribution.”
CORe50: a New Dataset and Benchmark for Continuous Object Recognition (2017, neutral): “In the ﬁrst group of datasets in Table 1 ( NORB , COIL-100 , iLAB20M , Washington RGB-D , BigBIRD , ALOI ), objects are positioned on turntables and acquisition is systematically controlled in term of pose/lighting.”
Joint Unsupervised Learning of Deep Representations and Image Clusters (2016, methodology): “We choose three different types of datasets for testing: COIL100, MNIST-test and UMist, and three types of clustering algorithms including SC-LS [3], N-Cuts [52] and AC-PIC [69] for comparison since their better performance among all the algorithms.”
Distributed optimization of deeply nested systems (2012, methodology): “Work funded in part by NSF CAREER award IIS–0754089.”
YouTube-BoundingBoxes: A Large High-Precision Human-Annotated Data Set for Object Detection in Video (2017, background): “The exceptional pace of progress in recent years on the tasks of object recognition and detection in still images was enabled by the creation of large-scale, publicly available data sets [9, 13, 14, 19, 31, 32, 33, 41, 51, 56].”
Recent Advances in Open Set Recognition: A Survey (2018, background): “COIL20 [137]: has a total of 1440 gray images from 20 objects (72 images each object).”
Tensor Ring Decomposition (2016, methodology): “In this section, the proposed TR algorithms are evaluated and compared with TT and CP decompositions on Columbia Object Image Libraries (COIL)-100 dataset [56] that contains 7200 color images of 100 objects (72 images per object) with different reﬂectance and complex geometric characteristics.”
Capped Lp-Norm Graph Embedding for Photo Clustering (2016, methodology): “Four benchmark data sets COIL20 [10], BinAlpha [3], USPS [8] and MNIST [9], one recent released multimedia data set YFCC100M [18] are used.”
Deep active object recognition by joint label and action prediction (2015, methodology): “They used the COIL-100 dataset for their experiments, which consists of 7200 images of 100 toy objects rotated in depth ( Nayar et al., 1996 ).”
VisDA: A Synthetic-to-Real Benchmark for Visual Domain Adaptation (2018, unknown): “COIL20 [11] 1,440 20 1 (tools) Office [16] 1,410 31 3 (office) Caltech [2] 1,123 10 1 (office) CAD-office [12] 775 20 1 (office) Cross-Dataset [22] 70,000+ 40 12 (mixed)”
Semi-Supervised Classification by Low Density Separation (2005, background): “In Coil20, the data are gray-scale images of 20 different objects taken from different angles, in steps of 5 degrees [15].”
Optimization Techniques for Semi-Supervised Support Vector Machines (2008, methodology): “The Text data set is defined using the classes mac and mswindows of the Newsgroup20 data set preprocessed as in Szummer and Jaakkola (2001). The Uspst set contains the test data part of the well-known USPS data on handwritten digit recognition.”
How Useful Is Photo-Realistic Rendering for Visual Learning? (2016, background): “From the early days of COIL-100 [16] the Corel Stock Photos and 15 Scenes datasets [17] up to and including newer datasets such as PASCAL VOC [5] and Imagenet [4], we have experienced bias: every sample of the world is biased in some way – viewpoint, lighting, etc.”
Clustering with Deep Learning: Taxonomy and New Methods (2018, methodology): “For each object, 72 images were taken with a 5 degrees distance (Nene et al., 1996).”