Learning Discriminative and Shareable Features for Scene Classification

Zuo, Zhen; Wang, Gang; Shuai, Bing; Zhao, Lifan; Yang, Qingxiong; Jiang, Xudong

doi:10.1007/978-3-319-10590-1_36

Cited by 73 publications

(53 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For the dataset of Places365-standard, we use step size of 150,000 to decrease the learning rate and the whole train process stops at 600,000 iterations. To speed the training process, we use the multi-GPU extension [52] of Caffe [53] toolbox for our CNN training 4 . For testing our models, we use the common 5 crops (4 corners and 1 center) and their horizontal flipping for each image at a single scale, thus having 10 crops in total for each image.…”

Section: A Large-scale Datasets and Implementation Detailsmentioning

confidence: 99%

“…S CENE recognition is a fundamental problem in computer vision, and has received increasing attention in the past few years [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11]. Scene recognition not only provides rich semantic information of global structure [12], but also yields meaningful context that facilitates other related vision tasks, such as object detection [13], [14], [15], event recognition [16], [17], [18], and action recognition [19], [20], [21].…”

mentioning

confidence: 99%

See 1 more Smart Citation

Knowledge Guided Disambiguation for Large-Scale Scene Classification With Multi-Resolution CNNs

Wang

Guo

Huang

et al. 2017

IEEE Trans. on Image Process.

136

View full text Add to dashboard Cite

Convolutional neural networks (CNNs) have made remarkable progress on scene recognition, partially due to these recent large-scale scene datasets, such as the Places and Places2. Scene categories are often defined by multi-level information, including local objects, global layout, and background environment, thus leading to large intra-class variations. In addition, with the increasing number of scene categories, label ambiguity has become another crucial issue in large-scale classification. This paper focuses on large-scale scene recognition and makes two major contributions to tackle these issues. First, we propose a multi-resolution CNN architecture that captures visual content and structure at multiple levels. The multi-resolution CNNs are composed of coarse resolution CNNs and fine resolution CNNs, which are complementary to each other. Second, we design two knowledge guided disambiguation techniques to deal with the problem of label ambiguity: 1) we exploit the knowledge from the confusion matrix computed on validation data to merge ambiguous classes into a super category and 2) we utilize the knowledge of extra networks to produce a soft label for each image. Then, the super categories or soft labels are employed to guide CNN training on the Places2. We conduct extensive experiments on three large-scale image datasets (ImageNet, Places, and Places2), demonstrating the effectiveness of our approach. Furthermore, our method takes part in two major scene recognition challenges, and achieves the second place at the Places2 challenge in ILSVRC 2015, and the first place at the LSUN challenge in CVPR 2016. Finally, we directly test the learned representations on other scene benchmarks, and obtain the new state-of-the-art results on the MIT Indoor67 (86.7%) and SUN397 (72.0%). We release the code and models at https://github.com/wanglimin/MRCNN-Scene-Recognition.

show abstract

Section: A Large-scale Datasets and Implementation Detailsmentioning

confidence: 99%

mentioning

confidence: 99%

Knowledge Guided Disambiguation for Large-Scale Scene Classification With Multi-Resolution CNNs

Wang

Guo

Huang

et al. 2017

IEEE Trans. on Image Process.

136

View full text Add to dashboard Cite

show abstract

“…We have re-labeled those images and will make the labels publicly available. As the MIT Indoor67 dataset is widely used in performance evaluation, e.g., [11], [4], [5], [6], [18], [19], [20], [21], [16], [22], [23], [24], [25], [26], [14], [27], the notion of 4 % annotation error will likely contribute to revising the results obtained so far on this dataset. We show that the proposed model outperforms the top-performing approach [11] in robustness to occlusion and aspect change.…”

Section: A Our Approach and Contributionsmentioning

confidence: 99%

“…In [11], for example, the authors achieve state-of-the-art place categorization performance by extracting CNN features from images. Several researchers, e.g., [22], [4] proposed concatenating the CNN features with additional features to increase the performance of the final descriptor. Authors of [19] propose visual words extraction by CNN patches sampled at multiple scales.…”

Section: Related Workmentioning

confidence: 99%

Part-based room categorization for household service robots

Peter

Mandeljc

Leonardis

et al. 2016

2016 IEEE International Conference on Robotics and Automation (ICRA)

View full text Add to dashboard Cite

A service robot that operates in a previouslyunseen home environment should be able to recognize the functionality of the rooms it visits, such as a living room, a bathroom, etc. We present a novel part-based model and an approach for room categorization using data obtained from a visual sensor. Images are represented with sets of unordered parts that are obtained by object-agnostic region proposals, and encoded using state-of-the-art image descriptor extractor -a convolutional neural network (CNN). An approach is proposed that learns category-specific discriminative parts for the partbased model. The proposed approach was compared to the state-of-the-art CNN trained specifically for place recognition. Experimental results show that the proposed approach outperforms the holistic CNN by being robust to image degradation, such as occlusions, modifications of image scaling, and aspect changes. In addition, we report non-negligible annotation errors and image duplicates in a popular dataset for place categorization and discuss annotation ambiguities.

show abstract

“…However, shadows can also cause complications in image processing and computer vision. They can degrade the performance of object recognition [1], image feature extraction [2], scene analysis [3] and face recognition [4]. It is easy for the human eye to distinguish shadows from objects, but identifying shadows by computer is a challenging research problem.…”

Section: Introductionmentioning

confidence: 99%

Shadow detection using double-threshold pulse coupled neural networks

Jiang

Sun

2016

2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

A novel double-threshold pulse coupled neural networks (DT-PCNN) is proposed and applied to shadow detection. It attempts to reduce the false detection of shadows in a single image where the hue and brightness of some non-shadow regions are similar to or even lower than those of shadows. Shadows whose intensity and hue fall in between those of the scene and objectives are often viewed as non-shadows by the single dynamic threshold of PCNN. Moreover, entities with similar or darker hue and intensity may be wrongly classified as shadows. To solve this problem, two different dynamic thresholds that iteratively alter are designed. The upper and lower limits of detecting shadows are determined respectively by a higher threshold that decreases iteratively and a lower one that increases iteratively. The detection result is obtained by a fusion of two detection components. Experimental results demonstrate that compared to other tested methods, the misclassifications are significantly reduced and the shadows are more accurately extracted.Index Terms-Shadow detection, double-threshold pulse coupled neural networks (DTPCNN)

show abstract

Learning Discriminative and Shareable Features for Scene Classification

Cited by 73 publications

References 36 publications

Knowledge Guided Disambiguation for Large-Scale Scene Classification With Multi-Resolution CNNs

Knowledge Guided Disambiguation for Large-Scale Scene Classification With Multi-Resolution CNNs

Part-based room categorization for household service robots

Shadow detection using double-threshold pulse coupled neural networks

Contact Info

Product

Resources

About