Gimme Signals: Discriminative signal encoding for multimodal activity recognition

13 Mar 2020  ·  Raphael Memmesheimer, Nick Theisen, Dietrich Paulus ·

We present a simple, yet effective and flexible method for action recognition supporting multiple sensor modalities. Multivariate signal sequences are encoded in an image and are then classified using a recently proposed EfficientNet CNN architecture. Our focus was to find an approach that generalizes well across different sensor modalities without specific adaptions while still achieving good results. We apply our method to 4 action recognition datasets containing skeleton sequences, inertial and motion capturing measurements as well as \wifi fingerprints that range up to 120 action classes. Our method defines the current best CNN-based approach on the NTU RGB+D 120 dataset, lifts the state of the art on the ARIL Wi-Fi dataset by +6.78%, improves the UTD-MHAD inertial baseline by +14.4%, the UTD-MHAD skeleton baseline by 1.13% and achieves 96.11% on the Simitate motion capturing data (80/20 split). We further demonstrate experiments on both, modality fusion on a signal level and signal reduction to prevent the representation from overloading.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Action Recognition NTU RGB+D 120 Gimme Signals (AIS) Accuracy (Cross-Subject) 71.59 # 8
Accuracy (Cross-Setup) 70.8 # 8
Skeleton Based Action Recognition NTU RGB+D 120 Gimme Signals (Skeleton, AIS) Accuracy (Cross-Subject) 70.8% # 31
Accuracy (Cross-Setup) 71.6% # 31
Multimodal Activity Recognition UTD-MHAD Gimme Signals (Skeleton, AIS) Accuracy (CS) 93.33 # 2