as my title says I am having trouble mapping face data from the Live Link app to a metahuman.
Here is what I have done so far:
Created a UE5 project (Film/Video & Live Event)
Imported a metahuman (custom-made) into the project
Added required plugins to project (Live Link, ArKit, Apple-etc)
Connected Live Link mobile application to local network
Set the metahuman's animation controller to the Live Link feed
Calibrated the Live Link data within the Live Link application
The problem I am having:
Parts of the face are not responding at all (ex. metahuman's right eyebrow does not respond to me lifting my left eyebrow).
The left corner of the mouth seems to be stuck (ex. when I try to open my mouth, all points respond except for the single point stays where it is).
The mapping/naming of facial components seems to be mirrored/off/labeled wrong (ex. if I was to wink my right eye this would result in my right eye closing and my right cheek pressing upward. On the metahuman, the left eye would blink and the right cheek would raise.)
These issues are very frustrating as I can not seem to get past this basic calibration. I see online people using these same tools and getting results with the metahuman's facial movements that are really clean. Is there something I am missing? I know that after the metahuman has been calibrated, I will create sequences, am I supposed to be modifying these value there? I am not sure... I have commented on every video I can find and I have posted this question in the Unreal Discord (here) with basically no help.
Note: I don't need a full solution, I just need to be pointed in the right direction! Please let me know if there is anything I am missing in my setup or calibration workflow.
thanks for reading.
Had the exact same problem with a customer project, and the issue was it wasn't using the correct animation blueprint. Once we switched it to the correct animation blueprint for LiveLink the issue was immediately resolved. I remember that exact same facial expression - sorry as I know this was a frustrating one for the customer as well.
It seemed that I was doing everything right. After upgrading from 5.0.3 to 5.1 the issue stopped completely.
Related
I m trying to setup a few image categorization tasks on Mechanical Turk sandbox developer version. When I try to view the HIT(the annotation image), it appears blank. I clicked on the 'Accept HIT' button but I still couldn't see anything.
In order to make sure that nothing was wrong with my project setup in particular, I signed in as a worker to accept HITS on other projects involving image categorization. I still continue to see a blank image in their categorization projects, where the image to be annotated is supposed to be displayed.
Can anyone help with this problem?Thanks.
Problem solved - it was a simple browser incompatibility problem.
I know you have already answered this for yourself, however for other requesters out there I think this may be useful.
I was developing HITs and I too was having issues view the HIT in Sandbox in Chrome and Firefox, I realized that it had something to do with the script being blocked by the browser and the way to fix this was to "unblock the content" - usually a shield icon in the URL bar.
When further developing my HIT I added information about how to see the HIT in the description box of the HIT so turkers could read the instructions and then work on the HIT - to be absolutely clear to the turkers, I added a "(READ DESCRIPTION)" in the title so turkers would know where to look.
Hope this helps!
I'm checking out DivShot for the first time, and have fallen at the first hurdle.
On launching the app, I get the page editing screen, as expected, with the Inspector, Components and Page Options tabs on the right hand side of the screen.
But on the left hand side, there is no Navigator. And I can find no reference to the Navigator anywhere on the screen (thinking it might be a collapsed area, or something).
Looking at the Intro video on the Divshot website, there do seem to be some minor differences (the lack of the Navigator notwithstanding), so I am wondering whether something has changed in the latest release...
In any case, I can see no way of add files/resources/folder structures other than new pages, because of the absence of the Navigator.
Can anyone shed any light?
Much appreciated :)
It sounds like you're using the now legacy version of Divshot. We recently launched Divshot 1.0 with file support and a completely revamped project structure.
Legacy: http://app.divshot.com
New Divshot: http://builder.divshot.com
It's on a separate instance so beta users have time to export their pages and finish anything up in the old version before upgrading. Since we moved from generic folders and pages to actual files we're offering a manual upgrade.
If you'd like to update you can just go to https://api.divshot.com/upgrade and follow the instructions there. We're pushing a notice to the old version right away to eliminate any confusion. Sorry about that!
I've been programming for a while, but just recently decided to start developing for Mac OS X. I feel like I've come to grips with the basics of Objective-C and Cocoa development over the past week. I'm planning on making graphics apps, and as such am currently in the process of learning how to control Quartz compositions through a Cocoa app. I went through the tutorial that apple offers (with the Mac Engravings composition), and was able to create that just fine. In order to make sure that I truly understood what I learned, I decided to create my own composition and link it to a slightly more complicated Cocoa application.
Essentially, I have a composition that loads a movie or image through a Movie Loader patch, at which point it applies various filters to the frames before outputting it. In my Cocoa app, I've written code (or rather copied and pasted from other apple examples) that lets a user pick a file using an NSOpenPanel object. The filepath of the file they pick gets placed in a text-box that I placed in the app's window using Interface Builder. I binded the value of said text-box to the "Movie_Location" key in my composition, which is a published input in the Movie Loader patch that I'm using. However, no matter what I do, movies and images aren't loaded into this composition no matter what I try. The only thing that gets displayed is the default image that I have saved in that input from Quartz Composer (or nothing if I leave it blank before publishing).
I've added a Clear Color patch to the composition and binded that to a colorwell in my UI, and that successfully changes the color in my display, so I know that the composition and my Cocoa app are communicating. I've spent numerous hours at this point trying to figure out what's going on, and I've just about given up. Does the Movie Loader have any weird behaviors that I'm not aware of, or is there something obvious that I seem to be missing? I'd really appreciate any help or advice from anybody.
Thanks for reading through this...
Best,
Sami
There are two things I can think of as reasons why it is doing this:
The file path isn't formatted incorrectly. Try checking backslashes, colons, etc.
The box isn't updating the value. Try literally clicking in the text field and hitting enter.
That's all I can think of without seeing your quartz composition and/or code.
EDIT:
Check the other continuous box, in the general properties.
I figured this out yesterday. spudwaffle's second idea is what was going on. If I were to type a filepath in and hit enter, it would work just fine. I got this to work properly by just removing the bind and instead using the setValue:keyInPath: function that a patch controller offers. That said, is there some way to force a text-box to update? I remember seeing a "continuously update" or something like that button within the bind sub-menu in the inspector, but my code didn't work with that checked either.
Thanks to those of you that tried to help me! I really appreciate it.
Best,
Sami
i've followed a tutorial in wich you were Guided trough the development of a little, actual simple XML Application on iPhone.
I followed the Tutorial and run into one problem: The UITableView is empty, alltough the NSMutabelArray is not empty. The curious thing is, that i am not able to find diferences in the code of the Tutorial and my programmed code.
I tried different suggestions and ideas from UITableView gives empty table, does not load data, but either the iPhone screen turned Black or nothing was changed.
So i put both Projects (the downloaded Tutorial and my own Project) at my FTP server for you ftp://homes.dnsalias.com
If i missed information that is necessary or helpful please tell me. I hope someone, is willing to explain my mistake.
Thank you a lot...
Ok, i have had some basic Problems with UINavigationControllers and TableViews.
For those who found this Thread on google or something, here is a, for me helpfull, youtube link: link text
I am developing an iPhone application where i want to display three image in each row on scroll view where i need to click action on each image like Photo album in iPhone. I am not getting any sample code.
Hoping for help
subodh
There's plenty of sample code out there, I found this after only basic googling. You want to search for "UIImageView Iphone". It's also worth mentioning that Apple's very own Developer Center is extremely well written, and will teach you everything you need to know about iPhone programming.
Generally it is frowned down upon to say to look more or read documentation, but you really haven't looked at all. Especially because of Apple's own resource that tells you how to do almost anything, especially something like this. It's not something you can pick up and bits and pieces of and expect to be successful with, it really should be learned starting from the beginning and moving forward. This is especially true if you've never programmed before or are unfamiliar with C/Objective-C.
Three20 has a photo browser that is open source and works similarly to the iPhone's photo browser with some nice code examples. The images come from an image source object that can relate them to images in your apps bundle or images on the web. Looks like its Google group is here. I think that to use images in your bundle you use a URL formed like: bundle://image-name.png and not the typical use of the main bundle to get a path to resource.