Classification-Specific Parts for Improving Fine-Grained Visual Categorization

@article{Korsch2019ClassificationSpecificPF,
  title={Classification-Specific Parts for Improving Fine-Grained Visual Categorization},
  author={Dimitri Korsch and P. Bodesheim and Joachim Denzler},
  journal={ArXiv},
  year={2019},
  volume={abs/1909.07075}
}
Fine-grained visual categorization is a classification task for distinguishing categories with high intra-class and small inter-class variance. [...] Key Method The subsequently detected parts are then not only selected by a-posteriori classification knowledge, but also have an intrinsic spatial extent that is determined automatically. This is in contrast to most part-based approaches and even to available ground-truth part annotations, which only provide point coordinates and no additional scale information…Expand
Automated Visual Large Scale Monitoring of Faunal Biodiversity
Abstract To observe biodiversity, the variety of plant and animal life in the world or in a particular habitat, human observers make the most common examinations, often assisted by technicalExpand
Exploiting Web Images for Fine-Grained Visual Recognition by Eliminating Noisy Samples and Utilizing Hard Ones
TLDR
This paper proposes a novel approach for removing irrelevant samples from real-world web images during training, while employing useful hard examples to update the network and demonstrates that this approach is far superior to current state-of-the-art web-supervised methods. Expand
Fine-Grained Adversarial Semi-supervised Learning
In this paper we exploit Semi-Supervised Learning (SSL) to increase the amount of training data to improve the performance of Fine-Grained Visual Categorization (FGVC). This problem has not beenExpand
Lightweight Filtering of Noisy Web Data: Augmenting Fine-grained Datasets with Selected Internet Images
TLDR
While noisy web images consistently improve classification accuracies, this work suggests a new lightweight filtering strategy to retain a fraction of the data such that high accuracies are achieved with a significantly smaller training dataset. Expand
Re-rank Coarse Classification with Local Region Enhanced Features for Fine-Grained Image Recognition
TLDR
A retrievalbased coarse-to-fine framework is proposed, where the TopN classification results are rerank by using the local region enhanced embedding features to improve the Top1 accuracy and to obtain the discriminative regions for distinguishing the fine-grained images. Expand
TransFG: A Transformer Architecture for Fine-grained Recognition
TLDR
A novel transformer-based framework TransFG is proposed where all raw attention weights of the transformer are integrated into an attention map for guiding the network to effectively and accurately select discriminative image patches and compute their relations. Expand
WikiChurches: A Fine-Grained Dataset of Architectural Styles with Real-World Challenges
TLDR
A novel dataset for architectural style classification, consisting of 9,485 images of church buildings, and provides 631 bounding box annotations of characteristic visual features for 139 churches from four major categories, to serve as a benchmark for various research fields. Expand
Data-driven Meta-set Based Fine-Grained Visual Classification
TLDR
This work proposes a data-driven meta-set based approach to deal with noisy web images for fine-grained recognition, which can alleviate the harmful effects caused by out-of-distribution noise and properly exploit the in-dist distribution noisy samples for training. Expand
Data-driven Meta-set Based Fine-Grained Visual Recognition
TLDR
This work proposes a data-driven meta-set based approach to deal with noisy web images for fine-grained recognition that can alleviate the harmful effects caused by out-of-distribution noise and properly exploit the in-dist distribution noisy samples for training. Expand
Fine-Grained Visual Classification with Efficient End-to-end Localization
TLDR
This work presents an efficient localization module that can be fused with a classification network in an end-to-end setup and is able to achieve competitive recognition performance. Expand
...
1
2
...

References

SHOWING 1-10 OF 24 REFERENCES
Neural Activation Constellations: Unsupervised Part Model Discovery with Convolutional Networks
TLDR
An approach is presented that is able to learn part models in a completely unsupervised manner, without part annotations and even without given bounding boxes during learning, to find constellations of neural activation patterns computed using convolutional neural networks. Expand
The caltech-ucsd birds200-2011 dataset
  • Tech. Rep. CNS-TR-2011-001, California Institute of Technology
  • 2011
The Unreasonable Effectiveness of Noisy Data for Fine-Grained Recognition
TLDR
This work introduces an alternative approach, leveraging free, noisy data from the web and simple, generic methods of recognition, and demonstrates its efficacy on four fine-grained datasets, greatly exceeding existing state of the art without the manual collection of even a single label. Expand
The Whole Is More Than Its Parts? From Explicit to Implicit Pose Normalization
TLDR
This paper presents and compares the explicit pose normalization approach, a generalized framework for the final global average and bilinear pooling and a visualization approach for understanding and analyzing predictions of the model to address limitations in pose handling in traditional CNNs. Expand
Unsupervised Part Mining for Fine-grained Image Classification
TLDR
This work proposes a fully unsupervised part mining (UPM) approach to localize the discriminative parts without even image-level annotations, which largely improves the fine-grained classification performance. Expand
Weakly Supervised Complementary Parts Models for Fine-Grained Image Classification From the Bottom Up
TLDR
This paper builds complementary parts models in a weakly supervised manner to retrieve information suppressed by dominant object parts detected by convolutional neural networks and builds a bi-directional long short-term memory (LSTM) network to fuze and encode the partial information of these complementary parts into a comprehensive feature for image classification. Expand
Which and How Many Regions to Gaze: Focus Discriminative Regions for Fine-Grained Visual Categorization
TLDR
Compared with state-of-the-art methods on two widely-used fine-grained visual categorization datasets, the M2DRL approach achieves the best categorization accuracy. Expand
In Defense of Active Part Selection for Fine-Grained Classification
TLDR
This work proposes a novel method that is required by active part selection and enables sequential part-based classification and the training time of this method does not increase with the amount of possible part features. Expand
Large Scale Fine-Grained Categorization and Domain-Specific Transfer Learning
TLDR
This work proposes a measure to estimate domain similarity via Earth Mover's Distance and demonstrates that transfer learning benefits from pre-training on a source domain that is similar to the target domain by this measure. Expand
The iNaturalist Species Classification and Detection Dataset
TLDR
The iNaturalist species classification and detection dataset, consisting of 859,000 images from over 5,000 different species of plants and animals, is presented, which features visually similar species, captured in a wide variety of situations, from all over the world. Expand
...
1
2
3
...