On the Pitfalls of Batch Normalization for End-to-End Video Learning: A Study on Surgical Workflow Analysis

Batch Normalization's (BN) unique property of depending on other samples in a batch is known to cause problems in several tasks, including sequential modeling. Yet, BN-related issues are hardly studied for long video understanding, despite the ubiquitous use of BN in CNNs for feature extraction. Especially in surgical workflow analysis, where the lack of pretrained feature extractors has lead to complex, multi-stage training pipelines, limited awareness of BN issues may have hidden the benefits of training CNNs and temporal models end to end. In this paper, we %present and analyze known as well as novel pitfalls of BN in video learning, including issues specific to online tasks such as a 'cheating' effect in anticipation. We observe that BN's properties create major obstacles for end-to-end learning. However, using BN-free backbones, even simple CNN-LSTMs beat state of the art in two surgical tasks by utilizing adequate end-to-end training strategies which maximize temporal context. We conclude that awareness of BN's pitfalls is crucial for effective end-to-end learning in surgical tasks. By reproducing results on natural-video datasets, we hope our insights will benefit other areas of video learning as well. Code: \url{https://gitlab.com/nct_tso_public/pitfalls_bn}.