LTX 2.3 - Realistic Version
This one was trained using actual crash test and dashcam footage. It creates a realistic crash and deployment, but it does suffer with the "structure" of the airbag. In that it can deploy pretty limp. I recommend pairing it with the "Puffy" version, but turning the weight down on the Puffy version to .1 to .15.
In fact, the two LTX 2.3 versions pair well together and should be used in conjunction, shifting the weight balance as desired.
LTX 2.3 Version
Best version by far. Prompt for sound and dialog otherwise you'll get music. Works best at .75 strength.
LTX2 Version
This one is pretty straight forward. I don't quite have a handle on how to get quality outputs with LTX2 yet, but the lora does work. It wasn't trained on the sound of an airbag deployment (tried that and it was a nightmare) so you'll have to prompt it. Otherwise it, for some reason, adds in a horror soundtrack.
The physical motion works great, and the airbags look good. It's versatile and FAR dynamic than the Wan 2.2 loras.
Best results at .30-.60 strength. Higher and the sound goes to shit.
EDIT: Version 2 works better at higher strengths. It also has a decent deployment sound without much work being done, but I would recommend prompting the sound.
Version 2
Version 2 adds more dynamic motion to the character as well as a much faster and consistent bag inflation. It does also work well with post-inflation starting images for airbag interaction. It's pretty versatile and you can have an airbag come out of pretty much anywhere, but it still struggles with front views if a steering wheel is not clearly visible.
Version 1
A lora to add airbag deployment to Wan 2.2. This is separate from my Airbag Hug model for 2.1 but can work in conjunction with it. I'll make a Airbag Hug 2.2 shortly.
The starting image should, of course, be some one in a car. You can technically do a portrait and the video will shift to them in a car. I don't know how well it works deploying airbags in other settings.
Description
FAQ
Comments (9)
would it be possible for you to upload safetensors instead of zips?
Why? The safe tensors are in the zip file.
I tried uploading the individual files in addition to the zip but it was giving me errors. Since these are to be used in conjunction I'd rather keep them paired.
Because none of the software available can retrieve metadata for the files if you upload a zip. When LoRA manager or Swarm or any custom script tries to get information about models from civitai via the API, it uses the hash of the safetensors file. If that hash is not found in the index, that means that no URL with metadata is available for that hash, so no information about the files can be retrieved.
Uploading zips essentially breaks the way all LoRA managing/organizing software works.
If you download a few models and manually rename everything and manually record all metadata, then there is no problem.
However, if you collect LoRAs and preserve them for archival purposes, using scripts to gather metadata via API is the standard procedure. All major generation software has options or extensions for populating metadata for LoRAs via API, so things like triggers and strength and the creator's username and the description all can be added to the collection.
In the worst case scenario, I won't be able to even identify a LoRA uploaded this way and thus it will never be used, and history will tend to forget it as it is not documented adequately for use.
The bottom line is that for many many people, the practice of uploading zips to civitAI breaks the flow of adding new LoRAs to their collection using standard procedures.
If I have a zip, unzip it, and the two files have sensible names, then I may manually find them and annotate them. This is the fall back option if none of the software can categorize the LoRA. I'm not incapable of manually editing files. It does involve numerous files and lots of manual work that should be unnecessary, but it can be done.
But if I search for the name of the file and find nothing, I'm SOL.
If the hash can't be found via API, your files become orphaned.
@leisure_suit_larry I'll see what I can do.
@leisure_suit_larry Give it like 5 minutes. High noise will be on this page. Low noise will be on a separate page.
Alright! Fixed! Let me know if there are any issues.
@BagGuyArt - I really appreciate you taking the time to do this! I am not the only one who uses tools that collect metadata via API so you have done lots of us a favor. Thanks!!!
@leisure_suit_larry It's my pleasure! What you said makes sense and ultimately it's not really that much more effort for me.