Voice exercise detection (VAD) is a vital part in varied functions resembling speech recognition, speaker identification, and hands-free communication methods. With the rising demand for personalised and context-aware applied sciences, the necessity for efficient personalised VAD methods has grow to be paramount. On this paper, we current a comparative evaluation of Personalised Voice Exercise Detection (PVAD) methods to evaluate their real-world effectiveness. We introduce a complete method to evaluate PVAD methods, incorporating varied efficiency metrics resembling frame-level and utterance-level error Charges and Onset Detection latency, alongside user-level evaluation. By way of in depth experimentation and analysis, we offer a radical understanding of strengths and limitations of varied PVAD variants. This paper advances the understanding of PVAD know-how by providing insights into its efficacy and viability in sensible functions utilizing a complete set of metrics.