Abstract: Highlights•Design a Bayesian-based autoregressive model to simulate the HVS predictive behavior.•Integrate visual attention and foveated masking into DCT-based JND estimation.•Resolve cross-domain operation issues in DCT-based JND estimation.
Loading