Facial features are classified or grouped to generate the unique identity of individual human faces. The quality of face images detected should be sufficient to guarantee an accurate detection response and reduced true positive rate to identify the original human face, which in turn provide high security in public gathering applications. Though efficient face detection was ensured, trade off occurred between true positive rate and computational complexity. To address the challenge of increasing the true positive rate and reduce the computational complexity, this paper proposes a novel technique named Robust Face Detection using Delaunay Triangle (RFD-DT). In this model, first apply Spectral Cluster for efficient face detection from images acquired using Faces94 dataset. Subsequently, gender detection for the detected face is performed by applying Delaunay Triangle to guess whether the given image is a male or female. Finally, age estimation is carried out by applying Wrinkle Textured Seed Point. Extensive experiments carried out on the Faces94 dataset have revealed the outstanding performance of the proposed RFD-DT technique when benchmarked with various well established high-tech schemes. The results obtained by RFD-DT witness a significant increase in accuracy by improving the true positive rate with minimized computational complexity when compared with the results produced by the other methods.