The difference is whether there is a changing velocity or not.
I'm going to assume that you're defining acceleration in that second statement, because I'm not sure if you are and "changing velocity" is literally what acceleration means. In any case, both acceleration and velocity are vectors, both have a direction, and so a person's velocity sure as hell can't be constant when they're going in circles. Ergo, acceleration. I mean that's what force is, mass times acceleration, so if you move and you can feel it you're accelerating. Earth has gravity that can more than cancel it out, but we can't say the same for rides.
Somebody smarter and with more energy than me can probably come up with a rough estimate of the g's being pulled in each picture (ignoring gravity).
In any case, both acceleration and velocity are vectors, both have a direction, and so a person's velocity sure as hell can't be constant when they're going in circles.
Well, you can if the space-time is curved right, that's what orbits are, but that't just a nitpick.
In case you aren't joking, I believe the relevant statement is that acceleration and "a change in velocity over time" are the same thing.
If you imagine driving a car forward in a straight line, pressing the gas will make you accelerate (velocity becomes more forward). Pressing the brake will also make you accelerate (velocity becomes less forward). Turning the steering wheel will also make you accelerate (velocity points more to the left/more to the right).
While I'm at it, you can do physics computations in a rotating frame of reference, but it produces some fictious forces, and gets really wacky quickly. An easy example is that anything far enough away from the axis of rotation is moving faster than the speed of light.
We know, but also neither are acceleration and the magnitude of acceleration the same. Acceleration is a vector - it has a direction just as velocity does. Here's the definition I just copied from Google:
Acceleration: the rate of change of velocity per unit of time.
And here's how you write that in math:
a=∆v/∆t.
If you want to know more about the relationships between position, velocity, and acceleration, take a calculus class. Isaac Newton literally invented it to solve problems like this
Rotation is acceleration towards the center with a velocity perpendicular to the centre. Using a frame of reference that rotates along with the object doesn't change what is physically happening to that object, it just affects the way you observe what's happening. A rotating frame of reference is itself accelerating and each of those frames of reference are accelerating.
We don't feel the Earth's rotation because gravity is accelerating our entire body and surroundings at the same rate, plus it's not just the spinning keeping us in equilibrium; the left over force holds us on the ground.
The other two feel different because it's the structures that provide the acceleration towards the centre, which then pushes on our bodies where it makes contact, and then the structure of our bodies pulls the rest and you can feel the forces of things wanting to move in the direction of inertia but being pulled around the circle instead.
If rotating frames of reference weren't accelerating, turning a car would feel no different from going straight.