Not at all. All operations on TREZOR require the user to enter a PIN and a one-time password (OTP) . The attacker would have to guess your PIN which is very difficult because the one-time password makes brute-force attack almost impossible.
Aha - so it does use a PIN although I am guessing that the PIN is being typed in on the PC (so is vulnerable to key logging).
I think if the PIN were to be entered into the device directly (i.e. no chance of key logging) then it would be a much better (although requiring at least a numeric keypad - a more expensive) solution.