implementing multiple executions of a deep neural network, the deep neural network comprising a hooking portion outside a trusted execution environment (TEE), wherein an operator in the hooking portion, when executed outside the TEE, invokes a corresponding execution operator in the TEE;

determining, on the basis of the invocation of the corresponding execution operator in the TEE by the operator in the hooking portion, a computation graph corresponding to the hooking portion during the multiple executions of the deep neural network; and

executing, on the basis of the computation graph, the execution operator corresponding to the hooking portion in the TEE during the execution of the deep neural network after the multiple executions;

wherein the executing, on the basis of the computation graph, the execution operator corresponding to the hooking portion in the TEE comprises:

transmitting, after a first operator of the hooking portion invokes an execution operator corresponding to the first operator, an output from each invoked execution operator to a next operator indicated by the computation graph in the TEE to invoke the next operator until an execution operator corresponding to the last operator of the hooking portion is invoked; and

transmitting an output from the invocation of the execution operator corresponding to the last operator to the last operator of the hooking portion as an output from the last operator.