Learning with privileged information via adversarial discriminative modality distillation

TL;DR

This paper introduces an adversarial discriminative modality distillation method to enable RGB-based models to leverage depth information during training, improving performance on RGB-D vision tasks without requiring depth data at test time.

Contribution

It proposes a novel adversarial training approach to distill depth information into RGB representations, eliminating the need for multiple losses or complex hyperparameter tuning.

Findings

Achieves state-of-the-art results on NYUD object classification.

Improves video action recognition on NTU RGB+D dataset.

Effective modality distillation without additional test-time modalities.

Abstract

Heterogeneous data modalities can provide complementary cues for several tasks, usually leading to more robust algorithms and better performance. However, while training data can be accurately collected to include a variety of sensory modalities, it is often the case that not all of them are available in real life (testing) scenarios, where a model has to be deployed. This raises the challenge of how to extract information from multimodal data in the training stage, in a form that can be exploited at test time, considering limitations such as noisy or missing modalities. This paper presents a new approach in this direction for RGB-D vision tasks, developed within the adversarial learning and privileged information frameworks. We consider the practical case of learning representations from depth and RGB videos, while relying only on RGB data at test time. We propose a new approach to train a hallucination network that learns to distill depth information via adversarial learning, resulting in a clean approach without several losses to balance or hyperparameters. We report state-of-the-art results on object classification on the NYUD dataset and video action recognition on the largest multimodal dataset available for this task, the NTU RGB+D, as well as on the Northwestern-UCLA.