CLCE: An Approach to Refining Cross-Entropy and Contrastive Learning for Optimized Learning Fusion