Recognition of Instrument-Tissue Interactions in Endoscopic Videos via Action Triplets (2020-07-10T00:00:00.000000Z)

TL;DR

This work introduces a new laparoscopic dataset, CholecT40, consisting of 40 videos from the public dataset Cholesc80 in which all frames have been annotated using 128 triplet classes and proposes a trainable 3D interaction space, which captures the associations between the triplet components.

Abstract

Recognition of surgical activity is an essential component to develop context-aware decision support for the operating room. In this work, we tackle the recognition of fine-grained activities, modeled as action triplets \(\langle instrument, verb, target \rangle \) representing the tool activity. To this end, we introduce a new laparoscopic dataset, CholecT40, consisting of 40 videos from the public dataset Cholec80 in which all frames have been annotated using 128 triplet classes. Furthermore, we present an approach to recognize these triplets directly from the video data. It relies on a module called class activation guide, which uses the instrument activation maps to guide the verb and target recognition. To model the recognition of multiple triplets in the same frame, we also propose a trainable 3D interaction space, which captures the associations between the triplet components. Finally, we demonstrate the significance of these contributions via several ablation studies and comparisons to baselines on CholecT40.

Authors

P. Mascagni

7 papers

N. Padoy

15 papers

D. Mutter

8 papers

References29 items

Real-time automatic surgical phase recognition in laparoscopic sigmoidectomy using the convolutional neural network-based deep learning approach

Multi-Task Recurrent Convolutional Network with Correlation Loss for Surgical Video Analysis

Learning to Detect Human-Object Interactions With Knowledge

Multitask Learning of Temporal Connectionism in Convolutional Networks using a Joint Distribution Loss Function to Simultaneously Identify Tools and Phase in Surgical Videos

TL;DR

Abstract

Authors

References29 items

Real-time automatic surgical phase recognition in laparoscopic sigmoidectomy using the convolutional neural network-based deep learning approach

Multi-Task Recurrent Convolutional Network with Correlation Loss for Surgical Video Analysis

Learning to Detect Human-Object Interactions With Knowledge

Multitask Learning of Temporal Connectionism in Convolutional Networks using a Joint Distribution Loss Function to Simultaneously Identify Tools and Phase in Surgical Videos

Segmenting and classifying activities in robot-assisted surgery with recurrent neural networks

Weakly supervised convolutional LSTM approach for tool tracking in laparoscopic videos

Learning Human-Object Interactions by Graph Parsing Neural Networks

DeepPhase: Surgical Phase Recognition in CATARACTS Videos

Temporal coherence-based self-supervised learning for laparoscopic workflow analysis

Surgical Activity Recognition in Robot-Assisted Radical Prostatectomy using Deep Learning

Scaling Human-Object Interaction Recognition Through Zero-Shot Learning

Detecting and Recognizing Human-Object Interactions

Learning to Detect Human-Object Interactions

Surgical data science for next-generation interventions

Recognizing Surgical Activities with Recurrent Neural Networks

System events: readily accessible features for surgical phase detection

Automatic data-driven real-time segmentation and recognition of surgical workflow

EndoNet: A Deep Architecture for Recognition Tasks on Laparoscopic Videos

LapOntoSPM: an ontology for laparoscopic surgeries and its application to surgical phase recognition

Data-driven spatio-temporal RGBD feature encoding for action recognition in operating rooms

Smoke detection in endoscopic surgery videos: a first step towards retrieval of semantic events

Video based activity recognition in trauma resuscitation

Modeling and Segmentation of Surgical Workflow from Laparoscopic Video

Acquisition of Process Descriptions from Surgical Interventions

Episode Classification for the Analysis of Tissue/Instrument Interaction with Multiple Visual Cues

We present a large endoscopic action triplet dataset, CholecT40, for this task

We develop a novel deep learning model that uses weak localization information from tool prediction to guide verb and target detection

We propose the ﬁrst approach to recognize surgical actions as triplets of 〈 instrument, verb, target 〉 directly from surgical videos

Surgical Action

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names