اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدCorrelated Input-Dependent Label Noise in Large-Scale Image Classification
98
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Large scale image classification datasets often contain noisy labels. We take a principled probabilistic approach to modelling input-dependent, also known as heteroscedastic, label noise in these datasets. We place a multivariate Normal distributed latent variable on the final hidden layer of a neural network classifier. The covariance matrix of this latent variable, models the aleatoric uncertainty due to label noise. We demonstrate that the learned covariance structure captures known sources of label noise between semantically similar and co-occurring classes. Compared to standard neural network training and other baselines, we show significantly improved accuracy on Imagenet ILSVRC 2012 79.3% (+2.6%), Imagenet-21k 47.0% (+1.1%) and JFT 64.7% (+1.6%). We set a new state-of-the-art result on WebVision 1.0 with 76.6% top-1 accuracy. These datasets range from over 1M to over 300M training examples and from 1k classes to more than 21k classes. Our method is simple to use, and we provide an implementation that is a drop-in replacement for the final fully-connected layer in a deep classifier.
قيم البحث
اقرأ أيضاً
Learning with the textit{instance-dependent} label noise is challenging, because it is hard to model such real-world noise. Note that there are psychological and physiological evidences showing that we humans perceive instances by decomposing them in
to parts. Annotators are therefore more likely to annotate instances based on the parts rather than the whole instances, where a wrong mapping from parts to classes may cause the instance-dependent label noise. Motivated by this human cognition, in this paper, we approximate the instance-dependent label noise by exploiting textit{part-dependent} label noise. Specifically, since instances can be approximately reconstructed by a combination of parts, we approximate the instance-dependent textit{transition matrix} for an instance by a combination of the transition matrices for the parts of the instance. The transition matrices for parts can be learned by exploiting anchor points (i.e., data points that belong to a specific class almost surely). Empirical evaluations on synthetic and real-world datasets demonstrate our method is superior to the state-of-the-art approaches for learning from the instance-dependent label noise.
Label noise is frequently observed in real-world large-scale datasets. The noise is introduced due to a variety of reasons; it is heterogeneous and feature-dependent. Most existing approaches to handling noisy labels fall into two categories: they ei
ther assume an ideal feature-independent noise, or remain heuristic without theoretical guarantees. In this paper, we propose to target a new family of feature-dependent label noise, which is much more general than commonly used i.i.d. label noise and encompasses a broad spectrum of noise patterns. Focusing on this general noise family, we propose a progressive label correction algorithm that iteratively corrects labels and refines the model. We provide theoretical guarantees showing that for a wide variety of (unknown) noise patterns, a classifier trained with this strategy converges to be consistent with the Bayes classifier. In experiments, our method outperforms SOTA baselines and is robust to various noise types and levels.
The classification accuracy of deep learning models depends not only on the size of their training sets, but also on the quality of their labels. In medical image classification, large-scale datasets are becoming abundant, but their labels will be no
isy when they are automatically extracted from radiology reports using natural language processing tools. Given that deep learning models can easily overfit these noisy-label samples, it is important to study training approaches that can handle label noise. In this paper, we adapt a state-of-the-art (SOTA) noisy-label multi-class training approach to learn a multi-label classifier for the dataset Chest X-ray14, which is a large scale dataset known to contain label noise in the training set. Given that this dataset also has label noise in the testing set, we propose a new theoretically sound method to estimate the performance of the model on a hidden clean testing data, given the result on the noisy testing data. Using our clean data performance estimation, we notice that the majority of label noise on Chest X-ray14 is present in the class No Finding, which is intuitively correct because this is the most likely class to contain one or more of the 14 diseases due to labelling mistakes.
Collecting large-scale data with clean labels for supervised training of neural networks is practically challenging. Although noisy labels are usually cheap to acquire, existing methods suffer a lot from label noise. This paper targets at the challen
ge of robust training at high label noise regimes. The key insight to achieve this goal is to wisely leverage a small trusted set to estimate exemplar weights and pseudo labels for noisy data in order to reuse them for supervised training. We present a holistic framework to train deep neural networks in a way that is highly invulnerable to label noise. Our method sets the new state of the art on various types of label noise and achieves excellent performance on large-scale datasets with real-world label noise. For instance, on CIFAR100 with a $40%$ uniform noise ratio and only 10 trusted labeled data per class, our method achieves $80.2{pm}0.3%$ classification accuracy, where the error rate is only $1.4%$ higher than a neural network trained without label noise. Moreover, increasing the noise ratio to $80%$, our method still maintains a high accuracy of $75.5{pm}0.2%$, compared to the previous best accuracy $48.2%$. Source code available: https://github.com/google-research/google-research/tree/master/ieg
We study how robust current ImageNet models are to distribution shifts arising from natural variations in datasets. Most research on robustness focuses on synthetic image perturbations (noise, simulated weather artifacts, adversarial examples, etc.),
which leaves open how robustness on synthetic distribution shift relates to distribution shift arising in real data. Informed by an evaluation of 204 ImageNet models in 213 different test conditions, we find that there is often little to no transfer of robustness from current synthetic to natural distribution shift. Moreover, most current techniques provide no robustness to the natural distribution shifts in our testbed. The main exception is training on larger and more diverse datasets, which in multiple cases increases robustness, but is still far from closing the performance gaps. Our results indicate that distribution shifts arising in real data are currently an open research problem. We provide our testbed and data as a resource for future work at https://modestyachts.github.io/imagenet-testbed/ .
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
