In addition to the comment above, are there any other applications of the ease function outside of animation? Off the top of my head I feel like these types of functions would be used for fading in or out music clips in video/music editing programs
x-fa19
I agree with knguyen0811; it seems like ease functions can be applied not only to animations, but also for instances where we want to fade in/out. I suppose the key point to remember about ease functions is, as the lecture slide mentions, "starting/stopping gently," which definitely helps add a more realistic touch to things! Another example where this might be useful is in UI menu designs, where if someone taps the screen to pull out a menu, the interaction will feel more realistic if there's a slow ease out for the menu.
tyleryath
Adding on to x-fa19's comment above, here's a proper visualization of UI animation curves: https://medium.com/@RobertGummesson/a-look-at-uiview-animation-curves-part-1-191d9e6de0ab This medium article shows the different curves and how they affect the animation, my personal favorite are the spring animations. It's really interesting to see that these curves can be generalized to do so many different things!
peterqiu1997
From this link (https://codepen.io/radialglo/post/understanding-the-intuition-of-easing), we can see some more examples of easing, especially near the end of the article. Easing is used in film editing/panning/zooming, like the Ken Burns effect (https://en.wikipedia.org/wiki/Ken_Burns_effect), as well as momentum scrolling and many animation applications. For audio, I agree that ease-in / ease-out would be useful for cross-fading tracks to get a smooth transition.
shivamparikh
This function very closely resembles a warped version of the sigmoid activation function very commonly used in Neural Networks. Just thought that was an interesting relation between two computer science disciplines!
GitMerlin
The fact that derivatives of H1(t) at t=0 and t=1 are 0 enables us to fade in/out smoothly in various applications.
One reasoning for using sigmoid function in neural network is that it provides a smooth transition as input values change. Interestingly, the derivatives of sigmoid function at -inf and at inf are 0.
Are there also nice property about H2 and H3?
In addition to the comment above, are there any other applications of the ease function outside of animation? Off the top of my head I feel like these types of functions would be used for fading in or out music clips in video/music editing programs
I agree with knguyen0811; it seems like ease functions can be applied not only to animations, but also for instances where we want to fade in/out. I suppose the key point to remember about ease functions is, as the lecture slide mentions, "starting/stopping gently," which definitely helps add a more realistic touch to things! Another example where this might be useful is in UI menu designs, where if someone taps the screen to pull out a menu, the interaction will feel more realistic if there's a slow ease out for the menu.
Adding on to x-fa19's comment above, here's a proper visualization of UI animation curves: https://medium.com/@RobertGummesson/a-look-at-uiview-animation-curves-part-1-191d9e6de0ab This medium article shows the different curves and how they affect the animation, my personal favorite are the spring animations. It's really interesting to see that these curves can be generalized to do so many different things!
From this link (https://codepen.io/radialglo/post/understanding-the-intuition-of-easing), we can see some more examples of easing, especially near the end of the article. Easing is used in film editing/panning/zooming, like the Ken Burns effect (https://en.wikipedia.org/wiki/Ken_Burns_effect), as well as momentum scrolling and many animation applications. For audio, I agree that ease-in / ease-out would be useful for cross-fading tracks to get a smooth transition.
This function very closely resembles a warped version of the sigmoid activation function very commonly used in Neural Networks. Just thought that was an interesting relation between two computer science disciplines!
The fact that derivatives of H1(t) at t=0 and t=1 are 0 enables us to fade in/out smoothly in various applications. One reasoning for using sigmoid function in neural network is that it provides a smooth transition as input values change. Interestingly, the derivatives of sigmoid function at -inf and at inf are 0.