Our goal is to provide the best and freely available AI tools to the audience. We always prefer opensource tools rather than Freemium or Paid tools. Mainstream media is flooded with Paid AI Tools. Past week was really good for Image and Video Generation models.
Lets discuss best of the best AI tools which are opensource and can boost your efficiency at work place. So that you might get promotion or a praise 🙂
1) MIDI: Multi Instance Diffusion for Single image to 3d Scene Generation
Really interesting development from Open Source Community. This is fascinating AI tool.
This model generates 3D realistic model of any given picture. This can be used in many industries as a starting point to generate models and then modify them. The big thing is this is completely free to use & you can download it locally and use it in your machine.
Below is the example; Only one single picture and Boom! You got the realistic model.
| Type | Completely Open Source |
| Use Case | Single Image to 3D Scene Generation |
| Cost | Free |
| Official Website Link | Click here for Official Website of MIDI |
| Hugging Face Model Link for Local Use | Click here to download MIDI Model Locally |
| Hugging Face Demo Link | Click here for Hugging Face Demo of MIDI |
| GitHub Link | Click here for Official GitHub of MIDI |
| Official Paper | Click here for Official Paper of MIDI |
2) LHM: Free AI tool that can animate any Image to do anything
This is one more miracle from Chinese company Alibaba Group. This model can animate any image to do any kind of activity. Just look at the below demo video you will get the idea.
What you need is the Full Body image of the person. A reference video of the activity or motion that you want to do with the image character. That’s it! Feed it to the model and model will animate the image to do the exact similar activities of the video
| Type | Opensource |
| Cost | Free to Use |
| Use Case | Animate any Image character to do motion activities as the Reference Video |
| Official Website of LHM | Click here for official website of LHM |
| Official GitHub Code of LHM | Click here for official GitHub Code of LHM |
| Hugging Face Demo of LHM | Click here for Official Hugging Face Demo of LHM |
3) TrajectoryCrafter
It basically changes the camera angle of any video. Just look at the Demos of this model you will have idea what I am talking. Fully opensource video editor. Just check the demo video which company released.
| Type | Completely Open Source |
| Use Case | Video Camera Angle Change |
| Cost | Free |
| Official Website Link | Click here for Official Website of TrajectoryCrafter |
| Hugging Face Link | Click here for TrajectoryCrafter Demo Hugging Face |
| GitHub Link | Click here for Official GitHub Link of TrajectoryCrafter |
| Official Paper Link | Click here for TrajectoryCrafter Official Paper |
4) Tight Inversion
This model is super impressive it can edit the images while keeping the details and facial structure intact.
As you can see in below image; which is generated using this Tight Inversion. The Input Image and The Last Image which is edited by Tight Inversion method. The facial features of a beard man is almost the similar but he is wearing a hat.
| Type | Completely Open Source |
| Use Case | Consistent Character Face Edit |
| Cost | Free |
| Official Website | Click here for Official Website of Tight Inversion |
| Demo Hugging Face | Click here for Hugging Face Demo of Tight Inversion |
| Official Paper | Click here for Official Paper of Tight Inversion |
5) Ernie 4.5 from Baidu
Talking about AI and missing China is not possible. Recently Chinese Company Baidu released their new model called Ernie 4.5 which is a Chat GPT rival.
As always it is free for all. But you need google translate to use this because the official website and the model itself is in Chinese. But It is no big deal
| Cost | Free |
| Official Website | Click here to Use Ernie 4.5 |
| Special | Everything is similar like Qwen 2.5 but it also can analyze the Videos |
6) AI VFX Generators
Now I have lots of AI Models which are free to use. If you are video editor then these Free Tools are only for you. You can create many VFX effects like Squish the Objects, Explode the Objects, Crush the Object and many more.
| Use Case | AI VFX Video Generator |
| Cost | Free |
| Link of all AI VFX Generator | Click here to Use all AI VFX Effects for Free |
7) Coming Soon – VACE – All in One AI Video Generator & Editor
Another free and opensource video generator as well as editing tool VACE. The team is saying they will release the model and the model code soon. You can track the official page down below for future updates.
You can check the demo video attached below. See the prompt is given with two images and the VACE generated the fully realistic video with both of the images
| Type | Completely Opensource |
| Use Case | All in one Video Editor & Generator |
| Cost | Free |
| Official Website | Click here for Official Page of VACE |
| Official Research Paper | Click here for Official Research Paper of VACE |
AI never sleeps. Past week was fantastic. We got lots of AI news and we have soo many tools right now that we are not able to display them in one single article. So, lets explore all free AI tools currently available.
8) SpatialLM: AI Model for Spatial (3D) Understanding
You have heard of object detector from image right? For example; if there is a person, a box, animal etc in the image then it will make a box around it and mark it as an object. Yes, this is the same thing but it detects objects in the video. Just look at the demo video below.
This model is made by Many Core Team. It is completely free to use and opensource. So, in simplest term this AI tool can help you detect objects in the video.
Any Business Ideas? Use this model for Blind People Glasses with Camera that records the live video and give them audio instructions
| Type | Opensource |
| Cost | Free to Use |
| Use Case | Object Detector in Video |
| Official Website | Click here for official website of SpatialLM |
| Official GitHub Code | Click here for official GitHub Code of SpatialLM |
| Hugging Face | Click here for Official Hugging Face Model |
9) Thera: AI that Converts Most Blurriest Images in to Super Resolution Images
As the title itself says this model will convert your most blurriest images in to super resolution images. This model is made by team of professionals Mathematical Modeling and Machine Learning, University of Zurich.
See the demo below.
| Type | Opensource |
| Cost | Free to Use |
| Use Case | Covert Blurry Images in to Clear/Super Resolution Images |
| Official Website Thera | Click here for official website of Thera |
| Official GitHub Code Thera | Click here for official GitHub Code of Thera |
| Hugging Face Demo Thera | Click here for Official Hugging Face Demo of Thera |
10) Consistent Character Editing by Google’s Gemini
Remember the era when Google was lagging behind every other AI companies. Their image generation and text generation was not even average.
The situation is change as of now. Because the new Gemini is Powerful like a Hell!
For example; just check the image on the right side.
- One prompt saying create a specific character
- The other prompt says insert the same character in a screenshot of Gameplay.
I mean look at the consistency and the details. You don’t need any paid subscription for image generation now.
Gemini will do the all work mostly.
| Use Case | Text to Image with Consistent Character Editing |
| Cost | Free |
| Official Link | Click here to Use Gemini |
11) StdGEN: Free AI tool that can generate 3D model from single image
| Type | Opensource |
| Cost | Free to Use |
| Use Case | 3D model generation from a Single Image |
| Official Website of StdGEN | Click here for official website of StdGEN |
| Official GitHub Code StdGEN | Click here for official GitHub Code of LHM |
| Hugging Face Model StdGEN | Click here for Official Hugging Face Demo of LHM |
| Hugging Face Demo of StdGEN | Click here to try Hugging Face Demo of StdGEN |
Coming Soon AI Tools…
AudioX: Free AI Model That converts Text prompt in to Audio Effects.
For example if you are video creator and you want specific kind of background for a scene lets say; Dog Barking with Industrial machines running, people discussing in subtle voice. You will find these kind of separate audios and merge them right?
But this model will make your work too easy. You just have to give prompt or text instructions what kind of audio you want and this model will generate similar kind of audio for you.
See the demo of Text prompt to Audio Generator.
Note: This model and the code of this model is not released yet but team is saying that they will release it soon. Keep visiting AudioX official website
NVIDIA’s + Google DeepMind + Disney Research Partnered for Physical AI Robots
I guess biggest partnership ever happening in AI world is combining The Best Chip Making Company with Best AI Network. I don’t know what Disney is doing in this?
Nvidia partnered with Disney Research and Google DeepMind to create Physical AI Robots.
What are Physical AI Robots?
These are the robots that does the work just like the humans; does all activities physically just like the normal human does not in the Robotic format.
The topic will become more complex and lengthy; for more information you can see the below video.
