ViSAGE @ NTIRE 2026 Challenge on Video Saliency Prediction

TL;DR

The paper introduces ViSAGE, a multi-expert ensemble framework for video saliency prediction, which achieved top rankings in the NTIRE 2026 Challenge and demonstrates strong generalization.

Contribution

It proposes a novel adaptive gated ensemble approach that leverages diverse inductive biases for improved video saliency prediction.

Findings

ViSAGE ranked first on two evaluation metrics.

It outperformed most competitors on remaining metrics.

The method demonstrated strong generalization ability.

Abstract

In this report, we present our champion solution for the NTIRE 2026 Challenge on Video Saliency Prediction held in conjunction with CVPR 2026. To exploit complementary inductive biases for video saliency, we propose Video Saliency with Adaptive Gated Experts (ViSAGE), a multi-expert ensemble framework. Each specialized decoder performs adaptive gating and modulation to refine spatio-temporal features. The complementary predictions from different experts are then fused at inference. ViSAGE thereby aggregates diverse inductive biases to capture complex spatio-temporal saliency cues in videos. On the Private Test set, ViSAGE ranked first on two out of four evaluation metrics, and outperformed most competing solutions on the other two metrics, demonstrating its effectiveness and generalization ability. Our code has been released at https://github.com/iLearn-Lab/CVPRW26-ViSAGE.