Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Open-World Semi-Supervised Learning

About

A fundamental limitation of applying semi-supervised learning in real-world settings is the assumption that unlabeled test data contains only classes previously encountered in the labeled training data. However, this assumption rarely holds for data in-the-wild, where instances belonging to novel classes may appear at testing time. Here, we introduce a novel open-world semi-supervised learning setting that formalizes the notion that novel classes may appear in the unlabeled test data. In this novel setting, the goal is to solve the class distribution mismatch between labeled and unlabeled data, where at the test time every input instance either needs to be classified into one of the existing classes or a new unseen class needs to be initialized. To tackle this challenging problem, we propose ORCA, an end-to-end deep learning approach that introduces uncertainty adaptive margin mechanism to circumvent the bias towards seen classes caused by learning discriminative features for seen classes faster than for the novel classes. In this way, ORCA reduces the gap between intra-class variance of seen with respect to novel classes. Experiments on image classification datasets and a single-cell annotation dataset demonstrate that ORCA consistently outperforms alternative baselines, achieving 25% improvement on seen and 96% improvement on novel classes of the ImageNet dataset.

Kaidi Cao, Maria Brbic, Jure Leskovec• 2021

Related benchmarks

TaskDatasetResultRank
Object DetectionLVIS v1.0 (val)
APbbox2.03
518
Generalized Category DiscoveryImageNet-100
All Accuracy81.8
138
Generalized Category DiscoveryCIFAR-100
Accuracy (All)74.2
133
Generalized Category DiscoveryStanford Cars
Accuracy (All)33.8
128
Generalized Category DiscoveryCUB
Accuracy (All)36.3
113
Generalized Category DiscoveryCIFAR-10
All Accuracy97.1
105
Generalized Category DiscoveryFGVC Aircraft
Accuracy (All)31.9
82
Generalized Category DiscoveryCUB-200 (test)
Overall Accuracy40.7
63
Object DetectionLVIS--
59
Generalized Category DiscoveryHerbarium19
Score (All Categories)25
47
Showing 10 of 64 rows

Other info

Code

Follow for update