Animal extinction and biodiversity loss are critical issues impacting our planet today. This paper presents an interdisciplinary approach that merges the power of deep learning with the urgent necessity of wildlife conservation to aid ecologists in their fight against animal extinction. Utilizing a labeled and unlabeled dataset obtained from camera trap images from the Hajar Mountains of the United Arab Emirates, we trained and evaluated six different pre-trained deep learning models to assess how well they can be applied in real-world scenarios. Our analysis yielded promising results in accuracy and generalizability, with Adjusted Rand Index (ARI) scores comparing model predictions exceeding 0.6 for all models. Despite the substantial challenge presented by the imbalanced nature of our dataset, our models successfully classified a wide array of species, with our best model (DenseNet121) averaging a weighted accuracy of 72.78% with an F1 score of 0.957. The robustness of our models was further validated by a t-SNE analysis, which revealed coherent clustering of high-dimensional data. We developed a Graphical User Interface (GUI) application to bring our technology to non-technical users, allowing ecologists to classify images easily and leverage the power of AI in their conservation efforts. This study is a stride forward in leveraging artificial intelligence to aid ecological conservation, demonstrating the potential for machine learning to provide practical, effective solutions in real-world scenarios.