This review discusses how the brain merges sensory information from different modalities (vision, touch, audition) to form a coherent and robust percept. It highlights two main strategies: sensory combination, which maximizes information from different modalities, and sensory integration, which reduces variance in sensory estimates to increase reliability. The brain uses prior knowledge to interpret ambiguous sensory signals, and the integration of information is often modeled using the Maximum Likelihood Estimation (MLE) approach. The MLE model assumes that the most reliable estimate is the one with the lowest variance, and that the integrated estimate is a weighted sum of individual estimates, with weights proportional to their inverse variances. The review also discusses how sensory integration is affected by factors such as signal reliability, correlation between signals, and the context of the task. It emphasizes that the brain uses both bottom-up and top-down processes to integrate sensory information, and that this integration is crucial for accurate perception and action. The review concludes that the brain integrates sensory information in a statistically optimal way, and that this process is influenced by prior knowledge, signal reliability, and the specific task at hand.This review discusses how the brain merges sensory information from different modalities (vision, touch, audition) to form a coherent and robust percept. It highlights two main strategies: sensory combination, which maximizes information from different modalities, and sensory integration, which reduces variance in sensory estimates to increase reliability. The brain uses prior knowledge to interpret ambiguous sensory signals, and the integration of information is often modeled using the Maximum Likelihood Estimation (MLE) approach. The MLE model assumes that the most reliable estimate is the one with the lowest variance, and that the integrated estimate is a weighted sum of individual estimates, with weights proportional to their inverse variances. The review also discusses how sensory integration is affected by factors such as signal reliability, correlation between signals, and the context of the task. It emphasizes that the brain uses both bottom-up and top-down processes to integrate sensory information, and that this integration is crucial for accurate perception and action. The review concludes that the brain integrates sensory information in a statistically optimal way, and that this process is influenced by prior knowledge, signal reliability, and the specific task at hand.