This paper presents a novel feature fusion-based deep learning model (called CASU2Net) for fault detection in offshore wind turbines. The proposed CASU2Net model benefits of a two-step early fusion to enrich features in the final stage. Moreover, since previous studies did not consider uncertainty while model developing and also predictions, we take advantage of Monte Carlo dropout (MC dropout) to enhance the certainty of the results. To design fault detection model, we use five sensors and a sliding window to exploit the inherent temporal information contained in the raw time-series data obtained from sensors. The proposed model uses the nonlinear relationships among multiple sensor variables and the temporal dependency of each sensor on others which considerably increases the performance of fault detection model. A 10-fold cross-validation approach is used to verify the generalization of the model and evaluate the classification metrics. To evaluate the performance of the model, simulated data from a benchmark floating offshore wind turbine (FOWT) with supervisory control and data acquisition (SCADA) are used. The results illustrate that the proposed model would accurately disclose and classify more than 99% of the faults. Moreover, it is generalizable and can be used to detect faults for different types of systems.