This paper presents our contributions to the ASVspoof 5 challenge, focusing on Track 1 for both closed and open conditions. In the closed condition, we evaluate various ResNet architectures to identify the most effective model for spoofing detection. For the open condition, we utilize a ResNet architecture in speaker verification called the ResNet-101-SV model that integrates Squeeze-and-Excitation (SE) layers and Attentive Statistics Pooling layer. The ResNet-101-SV model is pre-trained on VoxCeleb2 and fine-tuned on a diverse set of datasets to enhance robustness against spoofing attacks. Our best model in the open condition achieves a 3.32% Equal Error Rate (EER) on the ASVspoof 5 evaluation progress set and a 2.15% EER on the Wild dataset, demonstrating significant improvements in spoofing detection performance.