ALPINE: Adaptive Localization for Parameter- and Sample-Efficient Few-Shot Learning

Hugging Face Daily Papers Papers
few-shot-learning image-classification parameter-efficient lightweight-architecture adaptive-localization meta-learning

Summary

ALPINE introduces an ultra-lightweight spatial-relational architecture for few-shot image classification that achieves accuracy gains with fewer parameters, faster convergence, and better robustness compared to baselines like Prototypical Networks and MAML.

Few-shot learning research is predominantly evaluated on accuracy alone, with limited attention to the parameter and training-sample budgets required to reach that accuracy - a real constraint for practitioners without large-scale compute. We present an ultra-lightweight (22,249-34,917 parameter) spatial-relational architecture for few-shot image classification that combines fixed Gabor edge-energy guidance with a windowed, content-adaptive patch locator. Under a strictly matched, iso-episode-budget protocol (250 meta-training episodes, 5 canonical seeds, 600 evaluation episodes per seed), our architecture achieves 5-shot accuracy gains, consistent across all five seeds, over Prototypical Networks, Relation Networks, and MAML on both CIFAR-FS and MiniImageNet, while using 27-53% fewer parameters than any baseline. It also converges in fewer training episodes, generalizes better to an unseen fine-grained domain (CUB-200-2011 birds, zero retraining), and is more robust to 50% occlusion and 25% spatial translation than all three baselines. A series of falsification ablations - zeroing relational tokens at inference and retraining without them entirely - shows that the architecture's pairwise relational computation, while present, is not the primary driver of its performance; the content-adaptive patch locator is. We report this honestly, together with a capacity sweep showing a genuine accuracy plateau near 22-35k parameters, and release full seed-level results and checkpoint hashes for reproducibility.
Original Article
View Cached Full Text

Cached at: 09/23/26, 07:32 AM

Paper page - ALPINE: Adaptive Localization for Parameter- and Sample-Efficient Few-Shot Learning

Source: https://huggingface.co/papers/2609.22323

Abstract

Few-shotlearningresearchispredominantlyevaluatedonaccuracyalone,withlimitedattentiontotheparameterandtraining-samplebudgetsrequiredtoreachthataccuracy-arealconstraintforpractitionerswithoutlarge-scalecompute.Wepresentanultra-lightweight(22,249-34,917parameter)spatial-relationalarchitectureforfew-shotimageclassificationthatcombinesfixedGaboredge-energyguidancewithawindowed,content-adaptivepatchlocator.Underastrictlymatched,iso-episode-budgetprotocol(250meta-trainingepisodes,5canonicalseeds,600evaluationepisodesperseed),ourarchitectureachieves5-shotaccuracygains,consistentacrossallfiveseeds,overPrototypicalNetworks,RelationNetworks,andMAMLonbothCIFAR-FSandMiniImageNet,whileusing27-53%fewerparametersthananybaseline.Italsoconvergesinfewertrainingepisodes,generalizesbettertoanunseenfine-graineddomain(CUB-200-2011birds,zeroretraining),andismorerobustto50%occlusionand25%spatialtranslationthanallthreebaselines.Aseriesoffalsificationablations-zeroingrelationaltokensatinferenceandretrainingwithoutthementirely-showsthatthearchitecture’spairwiserelationalcomputation,whilepresent,isnottheprimarydriverofitsperformance;thecontent-adaptivepatchlocatoris.Wereportthishonestly,togetherwithacapacitysweepshowingagenuineaccuracyplateaunear22-35kparameters,andreleasefullseed-levelresultsandcheckpointhashesforreproducibility.

View arXiv pageView PDFGitHub0Add to collection

Models citing this paper1

#### NJ50/alpine-fewshot Image Classification• Updatedabout 2 hours ago • 1

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.22323 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.22323 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

FedOPAL: One-Shot Federated Learning via Analytic Visual Prompt Tuning

arXiv cs.AI

FedOPAL proposes a framework that adapts visual prompts as feature rectifiers for one-shot federated learning, achieving efficient gradient-free aggregation via analytic methods while outperforming existing analytical approaches and matching iterative methods with zero server-side training costs.

@mayfer: quick MNIST test

X AI KOLs Timeline

Sakana AI introduces PC-ALM, a local-learning alternative to backpropagation that enables training of 1000-layer neural networks using only local dynamics.