Fine-Grained Visual Classification of Aircraft
By Subhransu Maji, Esa Rahtu, Juho Kannala, Matthew B. Blaschko, Andrea Vedaldi. 2013
3,021 citations recorded by the source corpus.
This paper introduces FGVC-Aircraft, a new dataset containing 10,000 images of aircraft spanning 100 aircraft models, organised in a three-level hierarchy. At the finer level, differences between models are often subtle but always visually measurable, making visual recognition challenging but possible. A benchmark is obtained by defining corresponding classification tasks and evaluation protocols, and baseline results are presented. The construction of this dataset was made possible by the work of aircraft enthusiasts, a strategy that can extend to the study of number of other object classes. Compared to the domains usually considered in fine-grained visual classification (FGVC), for example animals, aircraft are rigid and hence less deformable. They, however, present other interesting modes of variation, including purpose, size, designation, structure, historical style, and branding.
Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Geology
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Aerospace Engineering
- Jitendra Malik
- Andrea Vedaldi
- Subhransu Maji
- Trevor Darrell
- Ross B. Girshick
- Yuanqing Lin
- Li Fei-Fei
- Cordelia Schmid
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Trevor Darrell
- Ming-Hsuan Yang
- Xiaogang Wang
- Bernt Schiele
- Jitendra Malik
- Yi Yang
In Geology
- Rahaf Aljundi
- Andrea Vedaldi
- Chen Change Loy
- Trevor Darrell
- Yonglong Tian
- L. V. D. van der Maaten
- Basura Fernando
- Andrew Zisserman
Citations in the corpus
- A Constructive Prediction of the Generalization Error Across Scales (2019, background): “Aircraft (Maji et al., 2013): 10K images of 100 different aircraft classes.”
- Do Better ImageNet Models Transfer Better? (2018, background): “These datasets covered a wide range of image classification tasks, including superordinate-level object classification (CIFAR-10 [43], CIFAR-100 [43], PASCAL VOC 2007 [22], Caltech-101 [24]); fine-grained object classification (Food-101 [5], Birdsnap [4], Stanford Cars [41], FGVC Aircraft [55], Oxfo”
- EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks (2019, neutral): “(Maji et al., 2013) 6,667 3,333 100 Oxford-IIIT Pets (Parkhi et al., 2012) 3,680 3,369 37 Food-101 (Bossard et al., 2014) 75,750 25,250 101”
- Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning (2020, methodology): “[92] Subhransu Maji, Esa Rahtu, Juho Kannala, Matthew B.”
- Fine-Grained Image Classification via Combining Vision and Language (2017, background): “And on the data side, more fine-grained domains have been covered, such as bird types [6, 7], dog species [8], plant breeds [9] , car types [10] and aircraft models [11].”
- Learning multiple visual domains with residual adapters (2017, background): “The decathlon challenge combines ten well-known datasets from multiple visual domains: FGVC-Aircraft Benchmark [24] contains 10,000 images of aircraft, with 100 images for each of 100 different aircraft model variants such as Boeing 737-400, Airbus A310.”
- Learning to Prompt for Vision-Language Models (2021, background): “%) on most fine-grained datasets including Flow-ers102, StanfordCars and FGVCAircraft, as well as on scene and action recognition datasets (i.e., SUN397 & UCF101).”
- CrossTransformers: spatially-aware few-shot transfer (2020, background): “In contrast, for test datasets like Aircraft [52], there is just a single corresponding class in ImageNet train; therefore, algorithms which don’t represent the intra-class variability for this class will be penalized.”
- AutoAugment: Learning Augmentation Policies from Data (2018, positive): “Even on datasets for which fine-tuning weights pre-trained on ImageNet does not help significantly [27], e.g. Stanford Cars [28] and FGVC Aircraft [29], training with the ImageNet policy reduces test set error by 1.16% and 1.76%, respectively.”
- Meta-Dataset: A Dataset of Datasets for Learning to Learn from Few Examples (2019, methodology): “Aircraft (Maji et al., 2013) A dataset of images of air-crafts spanning 102 model variants, with 100 images per class (Figure 2c).”
- Part-Based R-CNNs for Fine-Grained Category Detection (2014, background): “Recently, a large body of computer vision research has focused on the finegrained classification problem in a number of domains, such as animal breeds or species [16,26,31,33,37,46], plant species [1,2,5,34,35,39], and man-made objects [32,40].”
- The application of two-level attention models in deep convolutional neural network for fine-grained image classification (2014, background): “, classifying different bird types [22], dog breeds [11], flower species [15], aircraft models [14] etc.”