Towards General and Efficient Active Learning

Xie, Yichen; Tomizuka, Masayoshi; Zhan, Wei

doi:10.48550/arxiv.2112.07963

Cited by 3 publications

(8 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, DNNs require large-scale training data, while labeling such data is expensive and time-consuming. On the other hand, active learning (AL) selects the most suitable samples for annotation, which can boost the training of models using fewer annotations [Xie et al, 2021]. Thus, combining AL and DL can both reduce labeling cost and handle complex data.…”

Section: Introductionmentioning

confidence: 99%

A Comparative Survey of Deep Active Learning

Zhan¹,

Wang²,

Huang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Active Learning (AL) is a set of techniques for reducing labeling cost by sequentially selecting data samples from a large unlabeled data pool for labeling. Meanwhile, Deep Learning (DL) is datahungry, and the performance of DL models scales monotonically with more training data. Therefore, in recent years, Deep Active Learning (DAL) has risen as feasible solutions for maximizing model performance while minimizing the expensive labeling cost. Abundant methods have sprung up and literature reviews of DAL have been presented before. However, the performance comparison of different branches of DAL methods under various tasks is still insufficient and our work fills this gap. In this paper, we survey and categorize DAL-related work and construct comparative experiments across frequently used datasets and DAL algorithms. Additionally, we explore some factors (e.g., batch size, number of epochs in the training process) that influence the efficacy of DAL, which provides better references for researchers to design their own DAL experiments or carry out DAL-related applications. We construct a DAL toolkit, DeepAL + , by re-implementing many highly-cited DAL-related methods, and it will be released to the public.

show abstract

Section: Introductionmentioning

confidence: 99%

A Comparative Survey of Deep Active Learning

Zhan¹,

Wang²,

Huang³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…To solve this problem, Refs. [ 18 , 19 , 20 ] have been proposed to compute uncertainty algorithmically, using only task model learning. These methods do not require learning to predict uncertainty, making them easy to use in practice.…”

Section: Introductionmentioning

confidence: 99%

Non-Deep Active Learning for Deep Neural Networks

Kawano

Nota

Mochizuki

et al. 2022

Sensors

View full text Add to dashboard Cite

One way to improve annotation efficiency is active learning. The goal of active learning is to select images from many unlabeled images, where labeling will improve the accuracy of the machine learning model the most. To select the most informative unlabeled images, conventional methods use deep neural networks with a large number of computation nodes and long computation time, but we propose a non-deep neural network method that does not require any additional training for unlabeled image selection. The proposed method trains a task model on labeled images, and then the model predicts unlabeled images. Based on this prediction, an uncertainty indicator is generated for each unlabeled image. Images with a high uncertainty index are considered to have a high information content, and are selected for annotation. Our proposed method is based on a very simple and powerful idea: select samples near the decision boundary of the model. Experimental results on multiple datasets show that the proposed method achieves higher accuracy than conventional active learning methods on multiple tasks and up to 14 times faster execution time from 1.2 × 106 s to 8.3 × 104 s. The proposed method outperforms the current SoTA method by 1% accuracy on CIFAR-10.

show abstract

“…Nevertheless, for object detection tasks, each image may contain multiple candidate regions of target instances. Therefore, simply combining or averaging the uncertainties of instances or pixels does not adequately determine the overall image uncertainty [9]. To enhance the effectiveness of the active learning strategy on the detector's learning performance, it is essential to focus on the local feature maps of various prospective instance regions and eliminate background region interference.…”

mentioning

confidence: 99%

“…For example, MI-AOD [8] employs a pair of adversarial classifiers to align the representativeness and uncertainty of the target set, selecting images based on instance predictions. GEAL [9] uses feature point extraction for representative selection. CALD [28] employs a metric for data consistency in active learning.…”

mentioning

confidence: 99%

“…This can result in suboptimal outcomes when selecting representative samples. In contrast, GEAL [9] addresses representative selection by identifying clusters within images. Instead of computing posterior probabilities for each instance, it uses Transformerencoded instance-level intermediate features as the metric and leverages feature distance information for representative sampling.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Object Detection via Active Learning Strategy Based on Saliency of Local Features and Posterior Probability

Wang,

Liu,

Liu

2024

IEEE Access

View full text Add to dashboard Cite

Active learning is a strategy that involves the deliberate selection of training samples with significant discriminative features, to formulate a more efficient learning paradigm. While active learning has seen success in image classification, its application in object detection remains complex due to the multiple candidate object instances in an image and the interference of background regions on these instances. To tackle the challenge of sample selection in active learning for large object recognition datasets, this paper introduces a active learning method based on local feature saliency and posterior probability. Initially, the method selects active learning samples with discriminative saliency by clustering the multi-scale feature maps of the processed samples by self-attention weights. Subsequently, it calculates the posterior entropy based on the posterior probability of the current model's candidate instances. This is combined with the entropy encoding of feature points corresponding to instances in the cluster to further select uncertain samples in the object set. On the COCO dataset, our method demonstrates superior overall performance in the learning curve compared to existing object detection baselines. Moreover, under the same training configuration, the final detection accuracy shows a notable improvement of at least 2.36%, thus confirming the effectiveness of our method in enhancing learning and detection performance.

show abstract

Towards General and Efficient Active Learning

Cited by 3 publications

References 46 publications

A Comparative Survey of Deep Active Learning

A Comparative Survey of Deep Active Learning

Non-Deep Active Learning for Deep Neural Networks

Object Detection via Active Learning Strategy Based on Saliency of Local Features and Posterior Probability

Contact Info

Product

Resources

About