The development of a real-time monitoring system for meat freshness is crucial for ensuring food safety and enhancing the efficiency of the food supply chain. Herein, this study constructed a carboxymethyl cellulose-based colorimetric nanosensor array incorporating ionic liquid-tuned anthocyanins, a porous metal-organic framework particle (UiO-66), and nano-silica coating for real-time monitoring of pork freshness. Density functional theory calculations and multiple spectral analysis confirmed the mechanism by which ionic liquids tuned the anthocyanin color gamut. The sensor array's mechanical properties, thermal stability and sensitivity were significantly improved through the incorporation of UiO-66. The hydrophobic coating of the sensor array greatly enhanced the water contact angle rising from 76.5° to 134.8°, rendering excellent stability in high humidity conditions. The sensor array demonstrated good responsiveness and the durability up to 11 cycles to 5 common amine gases produced by meat spoilage. Additionally, the sensor array was integrated with 4 deep convolutional neural network models to predict pork freshness, all achieving accuracy rates above 97%. Notably, the VGG16 model achieved an accuracy rate of 99.16%. The designed monitoring system demonstrates significant potential for practical applications in intelligent meat packaging.