A Neural Network Model Can Explain Ventriloquism Aftereffect and Its Generalization across Sound Frequencies

Figure 8

Effects of network training with 𝐸^𝑎_0adapt = 20 and of network testing (after adaptation) with 𝐸^𝑎_0test = 20. The adaptation position was 80°, and the adaptation frequency was 1.1 kHz. The first column shows the untrained synapses that target neuron at azimuth 100° and frequency 𝐹; the second column shows the same synapses after the adaptation; the third column shows the response to the test stimulus applied at azimuth 80° and frequency 𝐹. The value for the frequency 𝐹 depends on the row of panels, as indicated inside the left panel in each row.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)