An official website of the United States government
Data augmentation is widely used for training a neural network given little labeled data. A common practice of augmentation training is applying a composition of multiple transformations sequentially to the data. Existing augmentation methods such as RandAugment randomly sample from a list of pre-selected transformations, while methods such as AutoAugment apply advanced search to optimize over an augmentation set of size $k^d$, which is the number of transformation sequences of length $$d$$, given a list of $$k$$ transformations. In this paper, we design efficient algorithms whose running time complexity is much faster than the worst-case complexity of $O(k^d)$, provably. We propose a new algorithm to search for a binary tree-structured composition of $$k$$ transformations, where each tree node corresponds to one transformation. The binary tree generalizes sequential augmentations, such as the SimCLR augmentation scheme for contrastive learning. Using a top-down, recursive search procedure, our algorithm achieves a runtime complexity of $O(2^d k)$, which is much faster than $O(k^d)$ as $$k$$ increases above $$2$$. We apply our algorithm to tackle data distributions with heterogeneous subpopulations by searching for one tree in each subpopulation and then learning a weighted combination, resulting in a \emph{forest} of trees. We validate our proposed algorithms on numerous graph and image datasets, including a multi-label graph classification dataset we collected. The dataset exhibits significant variations in the sizes of graphs and their average degrees, making it ideal for studying data augmentation. We show that our approach can reduce the computation cost by 43% over existing search methods while improving performance by 4.3%. The tree structures can be used to interpret the relative importance of each transformation, such as identifying the important transformations on small vs. large graphs.
more »
« less
Recurrent Networks for Guided Multi-Attention Classification
Attention-based image classification has gained increasing popularity in recent years. State-of-the-art methods for attention-based classification typically require a large training set and operate under the assumption that the label of an image depends solely on a single object (i.e., region of interest) in the image. However, in many real-world applications (e.g., medical imaging), it is very expensive to collect a large training set. Moreover, the label of each image is usually determined jointly by multiple regions of interest (ROIs). Fortunately, for such applications, it is often possible to collect the locations of the ROIs in each training image. In this paper, we study the problem of guided multi-attention classification, the goal of which is to achieve high accuracy under the dual constraints of (1) small sample size, and (2) multiple ROIs for each image. We propose a model, called Guided Attention Recurrent Network (GARN), for multi-attention classification. Different from existing attention-based methods, GARN utilizes guidance information regarding multiple ROIs thus allowing it to work well even when sample size is small. Empirical studies on three different visual tasks show that our guided attention approach can effectively boost model performance for multi-attention image classification.
more »
« less
- PAR ID:
- 10215765
- Date Published:
- Journal Name:
- The 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
- Page Range / eLocation ID:
- 412-420
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Multi-label classification (MLC), which assigns multiple labels to each instance, is crucial to domains from computer vision to text mining. Conventional methods for MLC require huge amounts of labeled data to capture complex dependencies between labels. However, such labeled datasets are expensive, or even impossible, to acquire. Worse yet, these pre-trained MLC models can only be used for the particular label set covered in the training data. Despite this severe limitation, few methods exist for expanding the set of labels predicted by pre-trained models. Instead, we acquire vast amounts of new labeled data and retrain a new model from scratch. Here, we propose combining the knowledge from multiple pre-trained models (teachers) to train a new student model that covers the union of the labels predicted by this set of teachers. This student supports a broader label set than any one of its teachers without using labeled data. We call this new problem knowledge amalgamation for multi-label classification. Our new method, Adaptive KNowledge Transfer (ANT), trains a student by learning from each teacher’s partial knowledge of label dependencies to infer the global dependencies between all labels across the teachers. We show that ANT succeeds in unifying label dependencies among teachers, outperforming five state-of-the-art methods on eight real-world datasets.more » « less
-
Multi-label classification is a challenging structured prediction task in which a set of output class labels are predicted for each input. Real-world datasets often have natural or latent taxonomic relationships between labels, making it desirable for models to employ label representations capable of capturing such taxonomies. Most existing multi-label classification methods do not do so, resulting in label predictions that are inconsistent with the taxonomic constraints, thus failing to accurately represent the fundamentals of problem setting. In this work, we introduce the multi-label box model (MBM), a multi-label classification method that combines the encoding power of neural networks with the inductive bias and probabilistic semantics of box embeddings (Vilnis, et al 2018). Box embeddings can be understood as trainable Venn-diagrams based on hyper-rectangles. Representing labels by boxes rather than vectors, MBM is able to capture taxonomic relations among labels. Furthermore, since box embeddings allow these relations to be learned by stochastic gradient descent from data, and to be read as calibrated conditional probabilities, our model is endowed with a high degree of interpretability. This interpretability also facilitates the injection of partial information about label-label relationships into model training, to further improve its consistency. We provide theoretical grounding for our method and show experimentally the model's ability to learn the true latent taxonomic structure from data. Through extensive empirical evaluations on both small and large-scale multi-label classification datasets, we show that BBM can significantly improve taxonomic consistency while preserving or surpassing the state-of-the-art predictive performance.more » « less
-
High-dimensional data is commonly encountered in various applications, including genomics, as well as image and video processing. Analyzing, computing, and visualizing such data pose significant challenges. Feature extraction methods are crucial in addressing these challenges by obtaining compressed representations that are suitable for analysis and downstream tasks. One effective technique along these lines is sparse coding, which involves representing data as a sparse linear combination of a set of exemplars. In this study, we propose a local sparse coding framework within the context of a classification problem. The objective is to predict the label of a given data point based on labeled training data. The primary optimization problem encourages the representation of each data point using nearby exemplars. We leverage the optimized sparse representation coefficients to predict the label of a test data point by assessing its similarity to the sparse representations of the training data. The proposed framework is computationally efficient and provides interpretable sparse representations. To illustrate the practicality of our proposed framework, we apply it to agriculture for the classification of crop diseases.more » « less
-
null (Ed.)Data collected from real-world environments often contain multiple objects, scenes, and activities. In comparison to single-label problems, where each data sample only defines one concept, multi-label problems allow the co-existence of multiple concepts. To exploit the rich semantic information in real-world data, multi-label classification has seen many applications in a variety of domains. The traditional approaches to multi-label problems tend to have the side effects of increased memory usage, slow model inference speed, and most importantly the under-utilization of the dependency across concepts. In this paper, we adopt multi-task learning to address these challenges. Multi-task learning treats the learning of each concept as a separate job, while at the same time leverages the shared representations among all tasks. We also propose a dynamic task balancing method to automatically adjust the task weight distribution by taking both sample-level and task-level learning complexities into consideration. Our framework is evaluated on a disaster video dataset and the performance is compared with several state-of-the-art multi-label and multi-task learning techniques. The results demonstrate the effectiveness and supremacy of our approach.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
