Since you specified the type of heatmap is not important, I have come up with the following solution.
import osmnx as ox
address_name='Melbourne'
#Import graph
G=ox.graph_from_address(address_name, distance=300)
#Make geodataframes from graph data
nodes, edges = ox.graph_to_gdfs(G, nodes=True, edges=True)
import numpy as np
#Create a new column in the nodes geodataframe with number of visits
#I have filled it up with random integers
nodes['visits'] = np.random.randint(0,1000, size=len(nodes))
#Now make the same graph, but this time from the geodataframes
#This will help retain the 'visits' columns
G = ox.save_load.gdfs_to_graph(nodes, edges)
#Then plot a graph where node size and node color are related to the number of visits
nc = ox.plot.get_node_colors_by_attr(G,'visits',num_bins = 5)
ox.plot_graph(G,fig_height=8,fig_width=8,node_size=nodes['visits'], node_color=nc)