Hey everyone, does anyone use Paraview here?

I am trying to generate a section view using Surface LIC with vector normalisation disabled. The effect I’m going for is in the linked figure on the left.

I uncheck the Normalize Vectors box and have played around with the number of steps and step size but I‘ve had no luck. Cheers!

https://www.paraview.org/Wiki/images/b/bc/Vec-norm.png

By the way , there is an extreme lack of videos on youtube about Signals and Systems subject , if you are good in this subject please consider making some videos on how to solve problems.

https://imgur.com/a/mKwuRuO - This is very poorly explained, can anyone help?

It says, looking at figure 13, we see that when....the integral extends from x-1 to 1. What? Makes zero sense to me. The ultimate "result" is obviously right, but where are these bounds coming from?

Trying to do practice problems from a FE study guide book... Problem is a cont. time convolution of the following functions: x(t) = 4(u(t)-u(t-4)) y(t) =2(u(t) -u(t-2)) where u(t) is the unit function.

Solutions in the book give the following 'regions' for computation of the integral...

Region 1 has integral limits of 0 to t for the region 0<t<2

Region 2 has integral limits of t-2 to t for the region 2<t<4

Region 3 has integral limits of t-2 to 4 for the region 4<t<6

I understand where the 'region' parts come from however I have no idea how they get the integral limits...

Essentially, how do they get from 0<t<2 to the integral limits of 0 to t? (same question for the other two regions)

I noticed a trend where if a piece wise function is given for h(t) the shifting function, the limits of integration for t < 0 is t -> 0, 0 < t < 1 is 0 -> t, and t > 1 is t - 1 -> t. Usually t < 0 is more than often zero so I'm not so sure about the limits on that integration.

Firstly, here’s the problem I’m trying to complete (c, not a)

When the problem is in the form

x(t) = f(t) + [0,t] ∫ k(t-τ)x(τ) dτ

I know how to solve that.

But in my case, k(t-τ) isn’t there. Does this mean I assume f(x) = e^(-t), k(x) = τ?

Then taking the Laplace of both sides I’d get..

ℒ{y} = (1/(s+1)) - ℒ{τ*y}

Not sure where to go from here

4th year biomedical engineering student here and I get the math behind it but I really have NO idea what it is or what is "happening" between the signals? Using this in a third class now and would be cool if someone could help me understand conceptually or what is happening, not just the math behind it.

EDIT: Trying to fix latex

I realized that I have a hole in my LTI knowledge, that while all systems represented by convolution integrals are LTI, I wasn't sure all LTI systems could be represented by convolution integrals. I know this is actually true, just curious about my proof.

Theorem: All continuous LTI systems can be represented as convolution integrals.

Proof: Let T be an LTI system operator, so y(t) = T[x(t)]. We know that $x(t) = \int_R x(\tau)h(t-\tau)d\tau$, so y(t) = T[x(t)] = T[$\int_R x(\tau)h(t-\tau)d\tau$], and $h(t)=T[\delta(t)]$.

Now, I am sort of hand-waving. I want to interchange T[] and the integral, but T isn't a limit, so this isn't a limit interchange. I think I'm going to justify the interchange by expressing the integral as a Riemann sum, and then using the assumption that T[] is linear. Is this valid?

So:

$$y(t) = T[\int x(\tau)h(t-\tau)d\tau]$$

$$T[\Sigma x(\tau) \delta(t-\tau) \Delta \tau]$$

$$\Sigma_{\Delta \tau -->0} T[x(\tau)\delta(t-\tau)\Delta \tau]$$

$$\Sigma_{\Delta \tau --> 0} x(\tau)h(t-\tau)d\tau$$

$$\int_R x(\tau)h(t-\tau)d\tau$$.

I understand that tau in the convolution integral of an continuous LTI system is just a variable of integration(dummy variable) that is used as an intermediate so that when you solve the integral your output is in terms of t (time). But if the integral is taken with respect to tau, is tau the same the same thing as t? I'm trying to think is tau a constant in h(t -tau) when the integral is evaluating t at each infinitesimal area within the bounds of integration? I don't know... but if anybody has a good explanation, it would be greatly appreciated. This is the integral I'm referring to https://drive.google.com/open?id=1aFUPn5JfAajP8Qr_ZzdLjkUsFN5VSTGF

In the new paper On the Integration of Self-Attention and Convolution, a research team from Tsinghua University, Huawei Technologies Ltd. and the Beijing Academy of Artificial Intelligence proposes ACmix, a mixed model that leverages the benefits of both self-attention and convolution for computer vision representation tasks while achieving minimum computational overhead compared to its pure convolution or self-attention counterparts.

Here is a quick read: Integrating Self-Attention and Convolution: Tsinghua, Huawei & BAAI’s ACmix Achieves SOTA Performance on CV Tasks With Minimum Cost.

The code and pretrained models will be released on the project’s GitHub. The paper On the Integration of Self-Attention and Convolution is on arXiv.

http://imgur.com/8NHwbKx

Hey guys, for this proof of the convolution integral, I understand what they mean by "assuming that the order of integration can be reversed," but why can you switch the integral bounds from T to infinity to 0 to t?

I don't understand what is going on in the image nor do I understand why they switched the bounds. Can somebody explain this to me?

I'm looking for a simple application problem involving line integrals. If it wasn't too associated with physics it would be great. I mean a little simple physics is no problem. Kind of a textbook example problem. I just learned about line integrals and need such a problem for a school project but can't think of anything smart. Thanks in advance!

https://imgur.com/a/vbjyma7

In the new paper On the Integration of Self-Attention and Convolution, a research team from Tsinghua University, Huawei Technologies Ltd. and the Beijing Academy of Artificial Intelligence proposes ACmix, a mixed model that leverages the benefits of both self-attention and convolution for computer vision representation tasks while achieving minimum computational overhead compared to its pure convolution or self-attention counterparts.

Here is a quick read: Integrating Self-Attention and Convolution: Tsinghua, Huawei & BAAI’s ACmix Achieves SOTA Performance on CV Tasks With Minimum Cost.

The code and pretrained models will be released on the project’s GitHub. The paper On the Integration of Self-Attention and Convolution is on arXiv.

In the new paper On the Integration of Self-Attention and Convolution, a research team from Tsinghua University, Huawei Technologies Ltd. and the Beijing Academy of Artificial Intelligence proposes ACmix, a mixed model that leverages the benefits of both self-attention and convolution for computer vision representation tasks while achieving minimum computational overhead compared to its pure convolution or self-attention counterparts.

Here is a quick read: Integrating Self-Attention and Convolution: Tsinghua, Huawei & BAAI’s ACmix Achieves SOTA Performance on CV Tasks With Minimum Cost.

The code and pretrained models will be released on the project’s GitHub. The paper On the Integration of Self-Attention and Convolution is on arXiv.