Self-exploration of movement possibilities and exploitation of natural dynamics are two crucial aspects of intelligent autonomous systems. We introduce a dynamical exploration strategy which combines chaotic neural activity with feedback-induced resonance. The underlying mechanism satisfies three conditions: (1) resonant patterns are discovered even in the presence of noisy sensory feedback; (2) no preprocessing of the feedback signal is required; and (3) exploration is an integral part of the mechanims. An embodied system endowed with our exploration strategy, can autonomously discover and tune into the intrinsic resonant modes most relevant to its body morphology. The mechanism not only allows to explore stable behavioral patterns, but also unstable ones. We demonstrate the role played by our exploration strategy on the emergence of movements through a simulation of a bipedal robot. We analyze with spectral and temporal methods the synchronicity of movements, spatio-temporal coordinations, and bifurcations in the couplings between neural, body, and environmental dynamics.