Childhood absence epilepsy is characterized by large and stereotyped 3 Hz spike-wave discharges recorded on scalp EEG often associated with brief episodes of lapse of consciousness. This study aims to detect the seizure onset, in childhood absence epilepsy, as early as possible. Interfering with absence seizures with sensory simulation has been shown to be possible on the condition that the stimulation occurs soon enough after the seizure onset, making early detection of spike-wave discharges a prerequisite for any viable closed-loop system of seizure prevention.We present four variations (two supervised, two unsupervised, two requiring a single spike-wave to mark a seizure, two requiring two spike-waves) of an algorithm designed to detect the onset of absence seizures from 4 scalp electrodes, and compare their performance with that of a state-of-the-art algorithm. We exploit the characteristic shape of spike-wave discharges to detect the seizure onset. Performance was assessed on clinical electroencephalograms from 63 patients with confirmed childhood absence epilepsy.The proposed approaches succeed in early detection of the seizure onset, contrary to the classical detection algorithm. They detect most seizures within a short delay of 200 (single spike-wave based algorithms) or 500 milliseconds (two spike-waves based algorithms). Perhaps due to the limits of their training, the performances of our supervised algorithms were comparable to those of our unsupervised ones.