Robust Character Labeling in Movie Videos: Data Resources and Self-supervised Feature Adaptation

TL;DR

This paper introduces a large-scale dataset and novel adaptation methods for robust face clustering in movies, addressing domain-specific challenges and improving face verification and clustering performance.

Contribution

It provides a new dataset of 169,000 face tracks from movies and proposes multiview correlation-based adaptation for better face embedding robustness.

Findings

Multiview correlation-based adaptation improves face embedding discriminability.

Weakly labeled data effectively enhances domain-specific feature adaptation.

The developed resources facilitate future research in character labeling in videos.

Abstract

Robust face clustering is a vital step in enabling computational understanding of visual character portrayal in media. Face clustering for long-form content is challenging because of variations in appearance and lack of supporting large-scale labeled data. Our work in this paper focuses on two key aspects of this problem: the lack of domain-specific training or benchmark datasets, and adapting face embeddings learned on web images to long-form content, specifically movies. First, we present a dataset of over 169,000 face tracks curated from 240 Hollywood movies with weak labels on whether a pair of face tracks belong to the same or a different character. We propose an offline algorithm based on nearest-neighbor search in the embedding space to mine hard-examples from these tracks. We then investigate triplet-loss and multiview correlation-based methods for adapting face embeddings to hard-examples. Our experimental results highlight the usefulness of weakly labeled data for domain-specific feature adaptation. Overall, we find that multiview correlation-based adaptation yields more discriminative and robust face embeddings. Its performance on downstream face verification and clustering tasks is comparable to that of the state-of-the-art results in this domain. We also present the SAIL-Movie Character Benchmark corpus developed to augment existing benchmarks. It consists of racially diverse actors and provides face-quality labels for subsequent error analysis. We hope that the large-scale datasets developed in this work can further advance automatic character labeling in videos. All resources are available freely at https://sail.usc.edu/~ccmi/multiface.