Multi-Resolution Weak Supervision for Sequential Data

Abstract

Since manually labeling training data is slow and expensive, recent industrial and scientific research efforts have turned to weaker or noisier forms of supervision sources. However, existing weak supervision approaches fail to model multi-resolution sources for sequential data, like video, that can assign labels to individual elements or collections of elements in a sequence. A key challenge in weak supervision is estimating the unknown accuracies and correlations of these sources without using labeled data. Multi-resolution sources exacerbate this challenge due to complex correlations and sample complexity that scales in the length of the sequence. We propose Dugong, the first framework to model multi-resolution weak supervision sources with complex correlations to assign probabilistic labels to training data. Theoretically, we prove that Dugong, under mild conditions, can uniquely recover the unobserved accuracy and correlation parameters and use parameter sharing to improve sample complexity. Our method assigns clinician-validated labels to population-scale biomedical video repositories, helping outperform traditional supervision by 36.8 F1 points and addressing a key use case where machine learning has been severely limited by the lack of expert labeled data. On average, Dugong improves over traditional supervision by 16.0 F1 points and existing weak supervision approaches by 24.2 F1 points across several video and sensor classification tasks.

arXiv commentNeurIPS 2019 (Conference on Neural Information Processing Systems)

References in corpus (9)

• Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
• YOLOv3: An Incremental Improvement
• Snorkel: Rapid Training Data Creation with Weak Supervision
• Identifiability of parameters in latent structure models with many observed variables
• Learning the Structure of Generative Models without Labeled Data
• Learning From Noisy Singly-labeled Data
• Large-scale, Fast and Accurate Shot Boundary Detection through Spatio-temporal Convolutional Neural Networks
• Learning to Learn from Weak Supervision by Full Supervision
• Rekall: Specifying Video Events using Compositions of Spatiotemporal Labels

Cited by in corpus (3)

• Ontology-driven weak supervision for clinical entity classification in electronic health records
• Fast and Three-rious: Speeding Up Weak Supervision with Triplet Methods
• Pairwise Feedback for Data Programming