Revisiting Surgical Action Triplet Recognition through the Lens of Robustness
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| This is a website made for Why Deep Surgical Models Fail?: Revisiting Surgical Action Triplet Recognition through the Lens of Robustness. |
| Surgical action triplet recognition provides a better understanding of the surgical scene. This task is of high relevance as it provides the surgeon with context-aware support and safety. The current go-to strategy for improving performance is the development of new network mechanisms. However, the performance of current state-of-the-art techniques is substantially lower than other surgical tasks. Why is this happening? This is the question that we address in this work. We present the first study to understand the failure of existing deep learning models through the lens of robustness and explainability. Firstly, we study current existing models under weak and strong δ−perturbations via an adversarial optimisation scheme. We then analyse the failure modes via feature based explanations. Our study reveals that the key to improving performance and increasing reliability is in the core and spurious attributes. Our work opens the door to more trustworthy and reliable deep learning models in surgical data science. |
 |
| We consider the current three SOTA techniques for our study: Tripnet, Attention Tripnet and Rendezvous; and extensively investigate the repercussion of deep features using four widely used backbones ResNet-18, ResNet-50, DenseNet-121 and Swin Transformer. Based on the networks output, we then evaluate robustness via adversarial optimisation to analyse the feature based explanations. |
 |
Yanqi Cheng, Lihao Liu, Shujun Wang, Yueming Jin, Carola-Bibiane Schönlieb and Angelica I. Aviles-Rivero. Why Deep Surgical Models Fail?: Revisiting Surgical Action Triplet Recognition through the Lens of Robustness. International Conference in Learning Representation (ICLR23) Workshop [Paper] [Bibtex] |
Acknowledgements |