RoboKA: KAN Informed Multimodal Learning for RoboCall Surveillance System

TL;DR

This paper introduces RoboKA, a multimodal learning framework using KAN for robocall detection, trained on a synthetic dataset Robo-SAr, and demonstrates superior performance over existing methods.

Contribution

The work presents Robo-SAr, a synthetic robocall dataset, and proposes RoboKA, a novel KAN-based multimodal fusion approach for robocall surveillance.

Findings

RoboKA outperforms unimodal and multimodal baselines in recall and F1-score.

RoboKA effectively models nonlinear interactions between acoustic and linguistic cues.

Synthetic dataset Robo-SAr enables robust robocall detection research.

Abstract

Wide exploration on robocall surveillance research is hindered due to limited access to public datasets, due to privacy concerns. In this work, we first curate Robo-SAr, a synthetic robocall dataset designed for robocall surveillance research. Robo-SAr comprises of ~200 unwanted and ~1200 legitimate synthetic robocall samples across three realistic adversarial axes: psycholinguistics-manipulated transcripts, emotion-eliciting speech, and cloned voices. We further propose RoboKA, a Kolmogorov-Arnold Network (KAN)-based multimodal fusion framework designed to model structured nonlinear interactions between acoustic and linguistic cues that characterize diverse adversarial robocall strategies. RoboKA first leverages cross-modal contrastive learning to align latent modality representations and feeds the resulting embeddings to a KAN-projection head for final classification. We benchmark RoboKA against strong unimodal and multimodal baselines in both in-domain and out-of-domain setups, finding RoboKA to surpass all baselines in terms of recall and F1-score.