Active learning in CV图像分类是计算机视觉、模式识别领域的研究热点,在智能交通、安全监控、机器人导航等领域有着广泛的应用。在图像分类中,需要大量有标记的样本来训练稳定的分类模型,以实现对未知图像的准确分类。但是在实际应用中,有标记的图像数量非常之少,无标记的图像却随处可见,且图像的人工标记是件费时费力的工作。为了减少人工标记工作量,主动学习(Active Learning)技术被引入到图像分类中。主动学习的主要思想是:在大量未标记的样本中,采用某种策略,挑选少量最有信息量且最具代表性的样本交给专家进行标记。使用标记过的样本训练模型,实现对未知样本的准确分类。主动学习的核心技术是如何设计准则来挑选最具信息量的样本,以最大程度提升分类模型的性能。
CVPR, pp.8753-8762, (2020)
The experiments on image classification and segmentation demonstrate that our model outperforms previous state-of-the-art methods and the initially sampling algorithm significantly improve the performance of our model
Cited by0BibtexViews132DOI
0
0
Samarth Sinha, Sayna Ebrahimi,Trevor Darrell
ICCV, pp.5971-5980, (2019)
We describe a pool-based semisupervised active learning algorithm that implicitly learns this sampling mechanism in an adversarial manner
Cited by84BibtexViews31DOI
0
0
CVPR, pp.9430-9440, (2019)
Our focus is on semantic segmentation; we believe that this work provides a highly promising research avenue towards other tasks in computer vision, including instance segmentation, object detection, activity understanding, or even visual-language embeddings
Cited by17BibtexViews76DOI
0
0
CVPR, (2019): 93-102
The loss prediction module is core to our task-agnostic active learning since it learns to imitate the loss defined in the target model
Cited by17BibtexViews108
1
0
2984303785, pp.3672-3680, (2019)
We have proposed an active learning method for object detectors based on convolutional neural networks
Cited by10BibtexViews28
0
0
Benjamin J. Meyer,Tom Drummond
ICRA, (2019): 2924-2931
We showed how a deep metric learning classification model is well suited to novelty detection and open set recognition
Cited by7BibtexViews17DOI
0
0
William H. Beluch,Tim Genewein,Andreas Nürnberger, Jan M. Köhler
CVPR, pp.9368-9377, (2018)
Through additional experiments we find that the difference in active learning performance can be explained by a combination of decreased model capacity and lower diversity of Monte Carlo Dropout ensembles
Cited by83BibtexViews40
0
0
ECCV, pp.212-229, (2018)
We introduced a novel active learning framework for temporal action localization
Cited by24BibtexViews33
0
0
arXiv: Learning, (2018)
That Adversarial Sampling for Active Learning outperforms random sampling on eight out of ten benchmarks
Cited by9BibtexViews49
0
0
International Journal of Computer Vision, no. 2 (2015): 113-127
We have proposed a new active learning algorithm Uncertainty Sampling with Diversity Maximization for visual concept recognition
Cited by261BibtexViews89DOI
0
0
IEEE Conference on Computer Vision and Pattern Recognition, (2015)
We present a variant of the expected model output change principle for active learning and discovery in the presence of unnameable instances
Cited by53BibtexViews10
0
0
ICCV, (2015): 2974-2982
In this paper we introduced an approach to exploiting the geometric priors inherent to images to increase the effectiveness of Active Learning for segmentation purposes
Cited by47BibtexViews24DOI
0
0
ICCV Workshops, (2015)
Occurrences of facial actions are sparse within this data, our active learning approach has allowed us to acquire hand-labeled positive examples from many different individuals up to 20x faster
Cited by43BibtexViews5DOI
0
0
IET Computer Vision, no. 3 (2015): 400-407
A new uncertain measure is developed from the multiple binary classifiers and the Gaussian kernel is applied for similarity measure
Cited by5BibtexViews7DOI
0
0
CVPR, pp.208-215, (2014)
This work takes a close look at active learning for relative attributes
Cited by56BibtexViews5DOI
0
0
CVPR, (2014): 564-571
We propose perplexity based graph construction and a new hierarchical subquery evaluation algorithm to combat this variability, and to release the potential of Expected Error Reduction
Cited by45BibtexViews6DOI
0
0
Xin Li,Yuhong Guo
CVPR, pp.859-866, (2013)
We presented a new adaptive active learning approach which combines an information density measure with a most uncertainty measure together in an adaptive way to conduct instance selection
Cited by178BibtexViews18DOI
0
0
computer vision and pattern recognition, pp.644-651, (2013)
We introduced a weighting scheme that intelligently reasons about the likelihood of any unlabeled image being a negative example for a category
Cited by102BibtexViews11DOI
0
0
ECCV, pp.86-98, (2008)
Our diverse feature set and accurate machine learning technique allow for precision which is superior to the state-of-the-art
Cited by149BibtexViews29DOI
0
0
ICCV, pp.1-8, (2007)
We have presented a discriminative probabilistic framework based on Gaussian Process priors and the Pyramid Match Kernel, and shown its utility for visual category recognition
Cited by342BibtexViews17DOI
0
0