Few-shot Tasks Research Articles

Geometric-perspective transfer learning for fast aerodynamic prediction in few-shot tasks

Geometric-perspective transfer learning for fast aerodynamic prediction in few-shot tasks

Semantic-Aware Dynamic Generation Networks for Few-Shot Human-Object Interaction Recognition.

Recognizing human-object interaction (HOI) aims at inferring various relationships between actions and objects. Although great progress in HOI has been made, the long-tail problem and combinatorial explosion problem are still practical challenges. To this end, we formulate HOI as a few-shot task to tackle both challenges and design a novel dynamic generation method to address this task. The proposed approach is called semantic-aware dynamic generation networks (SADG-Nets). Specifically, SADG-Net first assigns semantic-aware task representations for different batches of data, which further generates dynamic parameters. It obtains the features that highlight intercategory discriminability and intracategory commonality adaptively. In addition, we also design a dual semantic-aware encoder module (DSAE-Module), that is, verb-aware and noun-aware branches, to yield both action and object prototypes of HOI for each task space, which generalizes to novel combinations by transferring similarities among interactions. Extensive experimental results on two benchmark datasets, that is, humans interacting with common objects (HICO)-FS and trento universal HOI (TUHOI)-FS, illustrate that our SADG-Net achieves superior performance over state-of-the-art approaches, which proves its impressive effectiveness on few-shot HOI recognition.

Hybrid attentive prototypical network for few-shot action recognition

Most previous few-shot action recognition works tend to process video temporal and spatial features separately, resulting in insufficient extraction of comprehensive features. In this paper, a novel hybrid attentive prototypical network (HAPN) framework for few-shot action recognition is proposed. Distinguished by its joint processing of temporal and spatial information, the HAPN framework strategically manipulates these dimensions from feature extraction to the attention module, consequently enhancing its ability to perform action recognition tasks. Our framework utilizes the R(2+1)D backbone network, coupling the extraction of integrated temporal and spatial features to ensure a comprehensive understanding of video content. Additionally, our framework introduces the novel Residual Tri-dimensional Attention (ResTriDA) mechanism, specifically designed to augment feature information across the temporal, spatial, and channel dimensions. ResTriDA dynamically enhances crucial aspects of video features by amplifying significant channel-wise features for action distinction, accentuating spatial details vital for capturing the essence of actions within frames, and emphasizing temporal dynamics to capture movement over time. We further propose a prototypical attentive matching module (PAM) built on the concept of metric learning to resolve the overfitting issue common in few-shot tasks. We evaluate our HAPN framework on three classical few-shot action recognition datasets: Kinetics-100, UCF101, and HMDB51. The results indicate that our framework significantly outperformed state-of-the-art methods. Notably, the 1-shot task, demonstrated an increase of 9.8% in accuracy on UCF101 and improvements of 3.9% on HMDB51 and 12.4% on Kinetics-100. These gains confirm the robustness and effectiveness of our approach in leveraging limited data for precise action recognition.

Multi-task prompt tuning with soft context sharing for vision–language models

Vision–language models have recently shown great potential on many tasks in computer vision. Meanwhile, prior work demonstrates prompt tuning designed for vision–language models could acquire superior performance on few-shot image recognition compared to linear probe, a strong baseline. In practice, many few-shot tasks are inherently correlated, particularly within specialized domains. However, such information is overlooked previously. Inspired by the fact that modeling task relationship by multi-task learning can usually boost performance, we propose a novel method SoftCPT (Soft Context Sharing for Prompt Tuning) to tune pre-trained vision–language models on multiple target few-shot tasks jointly. Specifically, we design a task-shared meta network to generate prompt context for each task using task name together with a learnable task context as input. The parameters of this meta network as well as the task context are tuned on the joint training set of all tasks. As such, the prompt context of all tasks will be shared in a soft manner. Extensive experiments across four multi-task few-shot datasets covering 44 tasks and 1593 categories demonstrate that SoftCPT significantly outperforms single-task prompt tuning methods, highlighting the effectiveness of multi-task learning for vision–language prompt tuning.

Disentangled Feature Representation for Few-Shot Image Classification.

Learning the generalizable feature representation is critical to few-shot image classification. While recent works exploited task-specific feature embedding using meta-tasks for few-shot learning, they are limited in many challenging tasks as being distracted by the excursive features such as the background, domain, and style of the image samples. In this work, we propose a novel disentangled feature representation (DFR) framework, dubbed DFR, for few-shot learning applications. DFR can adaptively decouple the discriminative features that are modeled by the classification branch, from the class-irrelevant component of the variation branch. In general, most of the popular deep few-shot learning methods can be plugged in as the classification branch, thus DFR can boost their performance on various few-shot tasks. Furthermore, we propose a novel FS-DomainNet dataset based on DomainNet, for benchmarking the few-shot domain generalization (DG) tasks. We conducted extensive experiments to evaluate the proposed DFR on general, fine-grained, and cross-domain few-shot classification, as well as few-shot DG, using the corresponding four benchmarks, i.e., mini-ImageNet, tiered-ImageNet, Caltech-UCSD Birds 200-2011 (CUB), and the proposed FS-DomainNet. Thanks to the effective feature disentangling, the DFR-based few-shot classifiers achieved state-of-the-art results on all datasets.

Embedding enhancement with foreground feature alignment and primitive knowledge for few-shot learning

Few-Shot Learning (FSL) targets a model to quickly discriminate new categories with limited samples. While most methods struggle to effectively utilize the knowledge learned in the base classes, resulting in models that fail to alleviate the domain gap between training and evaluation. Conversely, humans often enable rapid discrimination by two ways: reinforcing impressions through comparing unknown categories with related objects in memory; Searching for the maximum shared information within the intra-class. Motivated by this, in this paper, we propose a novel Transformer-based Embedding Enhancement Network (TEEN) that adaptively leverages the knowledge learned from the base classes, which we refer to as ‘primitive knowledge’, to complete and distinguish the embeddings for novel classes. Additionally, to reduce interference from background features, we introduce the Transformer-based Foreground Feature Alignment (TFFA) to enhance the representation of image foreground. Ultimately, TEEN enhances the embeddings through foreground alignment and primitive knowledge completion, achieving improved performance in few-shot classification. Extensive experiments on inductive few-shot tasks demonstrate the effectiveness of our approach, achieving state-of-the-art results in cross-domain few-shot tasks.

The meta-learning method for the ensemble model based on situational meta-task.

The meta-learning methods have been widely used to solve the problem of few-shot learning. Generally, meta-learners are trained on a variety of tasks and then generalized to novel tasks. However, existing meta-learning methods do not consider the relationship between meta-tasks and novel tasks during the meta-training period, so that initial models of the meta-learner provide less useful meta-knowledge for the novel tasks. This leads to a weak generalization ability on novel tasks. Meanwhile, different initial models contain different meta-knowledge, which leads to certain differences in the learning effect of novel tasks during the meta-testing period. Therefore, this article puts forward a meta-optimization method based on situational meta-task construction and cooperation of multiple initial models. First, during the meta-training period, a method of constructing situational meta-task is proposed, and the selected candidate task sets provide more effective meta-knowledge for novel tasks. Then, during the meta-testing period, an ensemble model method based on meta-optimization is proposed to minimize the loss of inter-model cooperation in prediction, so that multiple models cooperation can realize the learning of novel tasks. The above-mentioned methods are applied to popular few-shot character datasets and image recognition datasets. Furthermore, the experiment results indicate that the proposed method achieves good effects in few-shot classification tasks. In future work, we will extend our methods to provide more generalized and useful meta-knowledge to the model during the meta-training period when the novel few-shot tasks are completely invisible.

Negatives Make a Positive: An Embarrassingly Simple Approach to Semi-Supervised Few-Shot Learning.

Semi-Supervised Few-Shot Learning (SSFSL) aims to train a classifier that can adapt to new tasks using limited labeled data and a fixed amount of unlabeled data. Various sophisticated methods have been proposed to tackle the challenges associated with this problem. In this paper, we present a simple but quite effective approach to predict accurate negative pseudo-labels of unlabeled data from an indirect learning perspective. We leverage these pseudo-labels to augment the support set, which is typically limited in few-shot tasks, e.g., 1-shot classification. In such label-constrained scenarios, our approach can offer highly accurate negative pseudo-labels. By iteratively excluding negative pseudo-labels one by one, we ultimately derive a positive pseudo-label for each unlabeled sample in our approach. The integration of negative and positive pseudo-labels complements the limited support set, resulting in significant accuracy improvements for SSFSL. Our approach can be implemented in just few lines of code by only using off-the-shelf operations, yet it outperforms state-of-the-art methods on four benchmark datasets. Furthermore, our approach exhibits good adaptability and generalization capabilities when used as a plug-and-play counterpart alongside existing SSFSL methods and when extended to generalized linear models.

Point2Real: Bridging the Gap between Point Cloud and Realistic Image for Open-World 3D Recognition

Recognition in open-world scenarios is an important and challenging field, where Vision-Language Pre-training paradigms have greatly impacted the 2D domain. This inspires a growing interest in introducing 2D pre-trained models, such as CLIP, into the 3D domain to enhance the ability of point cloud understanding. Considering the difference between discrete 3D point clouds and real-world 2D images, reducing the domain gap is crucial. Some recent works project point clouds onto a 2D plane to enable 3D zero-shot capabilities without training. However, this simplistic approach leads to an unclear or even distorted geometric structure, limiting the potential of 2D pre-trained models in 3D. To address the domain gap, we propose Point2Real, a training-free framework based on the realistic rendering technique to automate the transformation of the 3D point cloud domain into the Vision-Language domain. Specifically, Point2Real leverages a shape recovery module that devises an iterative ball-pivoting algorithm to convert point clouds into meshes, narrowing the gap in shape at first. To simulate photo-realistic images, a set of refined textures as candidates is applied for rendering, where the CLIP confidence is utilized to select the suitable one. Moreover, to tackle the viewpoint challenge, a heuristic multi-view adapter is implemented for feature aggregation, which exploits the depth surface as an effective indicator of view-specific discriminability for recognition. We conduct experiments on ModelNet10, ModelNet40, and ScanObjectNN datasets, and the results demonstrate that Point2Real outperforms other approaches in zero-shot and few-shot tasks by a large margin.

FeatWalk: Enhancing Few-Shot Classification through Local View Leveraging

Few-shot learning is a challenging task due to the limited availability of training samples. Recent few-shot learning studies with meta-learning and simple transfer learning methods have achieved promising performance. However, the feature extractor pre-trained with the upstream dataset may neglect the extraction of certain features which could be crucial for downstream tasks. In this study, inspired by the process of human learning in few-shot tasks, where humans not only observe the whole image (`global view') but also attend to various local image regions (`local view') for comprehensive understanding of detailed features, we propose a simple yet effective few-shot learning method called FeatWalk which can utilize the complementary nature of global and local views, therefore providing an intuitive and effective solution to the problem of insufficient local information extraction from the pre-trained feature extractor. Our method can be easily and flexibly combined with various existing methods, further enhancing few-shot learning performance. Extensive experiments on multiple benchmark datasets consistently demonstrate the effectiveness and versatility of our method.The source code is available at https://github.com/exceefind/FeatWalk.

Task Contamination: Language Models May Not Be Few-Shot Anymore

Large language models (LLMs) offer impressive performance in various zero-shot and few-shot tasks. However, their success in zero-shot or few-shot settings may be affected by task contamination, a potential limitation that has not been thoroughly examined. This paper investigates how zero-shot and few-shot performance of LLMs has changed chronologically over datasets released over time, and over LLMs released over time. Utilizing GPT-3 series models and several other recent open-sourced LLMs, and controlling for dataset difficulty, we find that datasets released prior to the LLM training data creation date perform surprisingly better than datasets released post the LLM training data creation date. This strongly indicates that, for many LLMs, there exists task contamination on zero-shot and few-shot evaluation for datasets prior to the LLMs' training data creation date. Additionally, we utilize training data inspection, training data extraction, and a membership inference attack, which reveal further evidence of task contamination. Importantly, we find that for tasks with no possibility of task contamination, LLMs rarely demonstrate statistically significant improvements over simple majority baselines, in both zero and few-shot settings.

Weak Distribution Detectors Lead to Stronger Generalizability of Vision-Language Prompt Tuning

We propose a generalized method for boosting the generalization ability of pre-trained vision-language models (VLMs) while fine-tuning on downstream few-shot tasks. The idea is realized by exploiting out-of-distribution (OOD) detection to predict whether a sample belongs to a base distribution or a novel distribution and then using the score generated by a dedicated competition based scoring function to fuse the zero-shot and few-shot classifier. The fused classifier is dynamic, which will bias towards the zero-shot classifier if a sample is more likely from the distribution pre-trained on, leading to improved base-to-novel generalization ability. Our method is performed only in test stage, which is applicable to boost existing methods without time-consuming re-training. Extensive experiments show that even weak distribution detectors can still improve VLMs' generalization ability. Specifically, with the help of OOD detectors, the harmonic mean of CoOp and ProGrad increase by 2.6 and 1.5 percentage points over 11 recognition datasets in the base-to-novel setting.

PMRC: Prompt-Based Machine Reading Comprehension for Few-Shot Named Entity Recognition

The prompt-based method has been proven effective in improving the performance of pre-trained language models (PLMs) on sentence-level few-shot tasks. However, when applying prompting to token-level tasks such as Named Entity Recognition (NER), specific templates need to be designed, and all possible segments of the input text need to be enumerated. These methods have high computational complexity in both training and inference processes, making them difficult to apply in real-world scenarios. To address these issues, we redefine the NER task as a Machine Reading Comprehension (MRC) task and incorporate prompting into the MRC framework. Specifically, we sequentially insert boundary markers for various entity types into the templates and use these markers as anchors during the inference process to differentiate entity types. In contrast to the traditional multi-turn question-answering extraction in the MRC framework, our method can extract all spans of entity types in one round. Furthermore, we propose word-based template and example-based template that enhance the MRC framework's perception of entity start and end positions while significantly reducing the manual effort required for template design. It is worth noting that in cross-domain scenarios, PMRC does not require redesigning the model architecture and can continue training by simply replacing the templates to recognize entity types in the target domain. Experimental results demonstrate that our approach outperforms state-of-the-art models in low-resource settings, achieving an average performance improvement of +5.2% in settings where access to source domain data is limited. Particularly, on the ATIS dataset with a large number of entity types and 10-shot setting, PMRC achieves a performance improvement of +15.7%. Moreover, our method achieves a decoding speed 40.56 times faster than the template-based cloze-style approach.

Auto-verbalizer filtering for prompt-based aspect category detection

Aspect category detection (ACD) is a basic task in sentiment analysis that aims to identify the specific aspect categories discussed in reviews. In the case of limited label resources, prompt-based models have shown promise in few-shot ACD. However, their current limitations lie in the manual selection or reliance on external knowledge for obtaining the verbalizer, a critical component of prompt learning that maps predicted words to final categories. To solve these issues, we propose an ACD method to automatically build the verbalizer in prompt learning. Our approach leverages the semantic expansion of category labels as prompts to automatically acquire initial verbalizer tokens. Additionally, we introduce an indicator mechanism for auto-verbalizer filtering to obtain reasonable verbalizer words and improve the predicting aspect category reliability of the method. In zero-shot task, our model exhibits an average performance improvement of 7.5% over the second-best model across four ACD datasets. For the other three few-shot tasks, the average performance improvement over the second-best model is approximately 2%. Notably, our method demonstrates effectiveness, particularly in handling general or miscellaneous category aspects.

Few-Shot Text Classification with an Efficient Prompt Tuning Method in Meta-Learning Framework

Meta-learning stands as a prevalent framework utilized in few-shot learning methods. Nonetheless, its efficacy hinges on substantial data availability during meta-training. Recent work adeptly tackled this hurdle by synergizing prompt tuning with the meta-learning paradigm, consequently attaining unparalleled performance on four benchmarks (FewRel, HuffPost, Reuters and Amazon). Nonetheless, the implementation efficacy of the previous method leaves room for enhancement, which is especially crucial when tuning larger language models. To this end, we introduce another expedited prompt tuning approach nested within the meta-learning framework. The novel approach normalizes the label information and sample information and uses the regression method to obtain the closed-form solution of each few-shot task, which significantly enhances inference speed, achieving a twofold improvement, while concurrently elevating average accuracy by [Formula: see text]% on the same benchmarks. Moreover, it demonstrates enhanced stability when faced with limited meta-training data, which is more applicable in many real scenarios where parallel data is rare. The source code is available to reproduce the results (http://github.com/Dr-Lv/EMPT).

Few-Shot Learning With a Strong Teacher.

Few-shot learning (FSL) aims to generate a classifier using limited labeled examples. Many existing works take the meta-learning approach, constructing a few-shot learner (a meta-model) that can learn from few-shot examples to generate a classifier. Typically, the few-shot learner is constructed or meta-trained by sampling multiple few-shot tasks in turn and optimizing the few-shot learner's performance in generating classifiers for those tasks. The performance is measured by how well the resulting classifiers classify the test (i.e., query) examples of those tasks. In this paper, we point out two potential weaknesses of this approach. First, the sampled query examples may not provide sufficient supervision for meta-training the few-shot learner. Second, the effectiveness of meta-learning diminishes sharply with the increasing number of shots (i.e., the number of training examples per class). To resolve these issues, we propose a novel meta-training objective for the few-shot learner, which is to encourage the few-shot learner to generate classifiers that perform like strong classifiers. Concretely, we associate each sampled few-shot task with a strong classifier, which is trained with ample labeled examples. The strong classifiers can be seen as the target classifiers that we hope the few-shot learner to generate given few-shot examples, and we use the strong classifiers to supervise the few-shot learner. We present an efficient way to construct the strong classifier, making our proposed objective an easily plug-and-play term to existing meta-learning based FSL methods. We validate our approach, (Learning with A Strong Teacher for few-SHOT learning), in combinations with many representative meta-learning methods. On several benchmark datasets including miniImageNet and tieredImageNet, our approach leads to a notable improvement across a variety of tasks. More importantly, with our approach, meta-learning based FSL methods can consistently outperform non-meta-learning based methods at different numbers of shots, even in many-shot settings, greatly strengthening their applicability.

Prototype Bayesian Meta-Learning for Few-Shot Image Classification.

Meta-learning aims to leverage prior knowledge from related tasks to enable a base learner to quickly adapt to new tasks with limited labeled samples. However, traditional meta-learning methods have limitations as they provide an optimal initialization for all new tasks, disregarding the inherent uncertainty induced by few-shot tasks and impeding task-specific self-adaptation initialization. In response to this challenge, this article proposes a novel probabilistic meta-learning approach called prototype Bayesian meta-learning (PBML). PBML focuses on meta-learning variational posteriors within a Bayesian framework, guided by prototype-conditioned prior information. Specifically, to capture model uncertainty, PBML treats both meta-and task-specific parameters as random variables and integrates their posterior estimates into hierarchical Bayesian modeling through variational inference (VI). During model inference, PBML employs Laplacian estimation to approximate the integral term over the likelihood loss, deriving a rigorous upper-bound for generalization errors. To enhance the model's expressiveness and enable task-specific adaptive initialization, PBML proposes a data-driven approach to model the task-specific variational posteriors. This is achieved by designing a generative model structure that incorporates prototype-conditioned task-dependent priors into the random generation of task-specific variational posteriors. Additionally, by performing latent embedding optimization, PBML decouples the gradient-based meta-learning from the high-dimensional variational parameter space. Experimental results on benchmark datasets for few-shot image classification illustrate that PBML attains state-of-the-art or competitive performance when compared to other related works. Versatility studies demonstrate the adaptability and applicability of PBML in addressing diverse and challenging few-shot tasks. Furthermore, ablation studies validate the performance gains attributed to the inference and model components.

Enhancing Few-Shot CLIP With Semantic-Aware Fine-Tuning.

Learning generalized representations from limited training samples is crucial for applying deep neural networks in low-resource scenarios. Recently, methods based on contrastive language-image pretraining (CLIP) have exhibited promising performance in few-shot adaptation tasks. To avoid catastrophic forgetting and overfitting caused by few-shot fine-tuning, existing works usually freeze the parameters of CLIP pretrained on large-scale datasets, overlooking the possibility that some parameters might not be suitable for downstream tasks. To this end, we revisit CLIP's visual encoder with a specific focus on its distinctive attention pooling layer, which performs a spatial weighted-sum of the dense feature maps. Given that dense feature maps contain meaningful semantic information, and different semantics hold varying importance for diverse downstream tasks (such as prioritizing semantics like ears and eyes in pet classification tasks rather than side mirrors), using the same weighted-sum operation for dense features across different few-shot tasks might not be appropriate. Hence, we propose fine-tuning the parameters of the attention pooling layer during the training process to encourage the model to focus on task-specific semantics. In the inference process, we perform residual blending between the features pooled by the fine-tuned and the original attention pooling layers to incorporate both the few-shot knowledge and the pretrained CLIP's prior knowledge. We term this method as semantic-aware fine-tuning (). is effective in enhancing the conventional few-shot CLIP and is compatible with the existing adapter approach (termed ). Extensive experiments on 11 benchmarks demonstrate that both and significantly outperform the second-best method by + 1.51 % and + 2.38 % in the one-shot setting and by + 0.48 % and + 1.37 % in the four-shot setting, respectively.

Pseudo-Labeling Based Practical Semi-Supervised Meta-Training for Few-Shot Learning.

Most existing few-shot learning (FSL) methods require a large amount of labeled data in meta-training, which is a major limit. To reduce the requirement of labels, a semi-supervised meta-training (SSMT) setting has been proposed for FSL, which includes only a few labeled samples and numbers of unlabeled samples in base classes. However, existing methods under this setting require class-aware sample selection from the unlabeled set, which violates the assumption of unlabeled set. In this paper, we propose a practical semi-supervised meta-training setting with truly unlabeled data to facilitate the applications of FSL in realistic scenarios. To better utilize both the labeled and truly unlabeled data, we propose a simple and effective meta-training framework, called pseudo-labeling based meta-learning (PLML). Firstly, we train a classifier via common semi-supervised learning (SSL) and use it to obtain the pseudo-labels of unlabeled data. Then we build few-shot tasks from labeled and pseudo-labeled data and design a novel finetuning method with feature smoothing and noise suppression to better learn the FSL model from noise labels. Surprisingly, through extensive experiments across two FSL datasets, we find that this simple meta-training framework effectively prevents the performance degradation of various FSL models under limited labeled data, and also significantly outperforms the representative SSMT models. Besides, benefiting from meta-training, our method also improves several representative SSL algorithms as well. We provide the training code and usage examples at https://github.com/ouyangtianran/PLML.

Generalizable and Discriminative Representations for Adversarially Robust Few-Shot Learning.

Few-shot image classification (FSIC) is beneficial for a variety of real-world scenarios, aiming to construct a recognition system with limited training data. In this article, we extend the original FSIC task by incorporating defense against malicious adversarial examples. This can be an arduous challenge because numerous deep learning-based approaches remain susceptible to adversarial examples, even when trained with ample amounts of data. Previous studies on this problem have predominantly concentrated on the meta-learning framework, which involves sampling numerous few-shot tasks during the training stage. In contrast, we propose a straightforward but effective baseline via learning robust and discriminative representations without tedious meta-task sampling, which can further be generalized to unforeseen adversarial FSIC tasks. Specifically, we introduce an adversarial-aware (AA) mechanism that exploits feature-level distinctions between the legitimate and the adversarial domains to provide supplementary supervision. Moreover, we design a novel adversarial reweighting training strategy to ameliorate the imbalance among adversarial examples. To further enhance the adversarial robustness without compromising discriminative features, we propose the cyclic feature purifier during the postprocessing projection, which can reduce the interference of unforeseen adversarial examples. Furthermore, our method can obtain robust feature embeddings that maintain superior transferability, even when facing cross-domain adversarial examples. Extensive experiments and systematic analyses demonstrate that our method achieves state-of-the-art robustness as well as natural performance among adversarially robust FSIC algorithms on three standard benchmarks by a substantial margin.