This work presents an energy-aware scheduling algorithm for deep neural network (DNN) inference tasks under energy, memory, and deadline constraints. By leveraging a precomputed lookup table (LUT) of performance metrics, the scheduler dynamically balances energy consumption and execution time through optimized batch sizes, execution modes, and GPU frequency adjustments. Sequential execution minimizes energy, while concurrent execution is used only when necessary to meet deadlines. Real-time regrouping and resource reallocation further enhance efficiency. Experiments on CIFAR-10 and CIFAR-100 validate the scheduler's ability to meet deadlines while minimizing energy consumption, demonstrating its utility for time-critical, energy-sensitive applications.