Source code and realases: AIGod project
AIGod - program to recognize objects in images, videos, webcams. The program has a graphical menu to facilitate users' use.
- the choice of a specific model, which is trained to identify certain objects of a given topic. Each model defines objects fairly precisely 70-80%, but I will improve the discovery of objects to 100% accuracy
- select certain objects from all in the model for the recognition task
- display the result and automatically save it to a folder
- selection of the language of the program
- the ability to check the program update
- comparison of people’s faces in the selected database
- object counting
- track the movement of objects
- sound notification about the detection of objects
- description of Wikipedia objects found
- the ability to select a program to view the result
- the ability to choose the device through which the analysis will be carried out
- recognition history
- choice of visual design
- and much more
- Transport
- Household objects
- People
- Animals
- Sport
- Plants
- Food
- Closes
- Art
- Smoke (Fire)
- Car license plates
- Deepfake and other
- Colors
- Movement of objects
- Emotions
- Other
Help in the development of the project. If you have another currency - let me know.
Artificial intelligence of recognition: The program uses the Yolov5 recognition model. The program is written in Python.
How fast is object recognition, how many resources are required: The program requires a lot of computer resources for fast work. Object recognition takes place instantly.
Project status: The project may be slow, but I will always try to answer questions.
Application pros:
- Free
- Open source code
- Graphic window
- Wide range of features and options
- Relevance and activity of the project
- Large range of models for object recognition
- Low weight models and fast speed of their work
- Multilingualism
- Cross-platform (Windows/Linux)
Application cons:
- The speed is lower than that of narrowly focused competitors
- Weight of the whole project
- The results of the models are not always accurate (but it is possible for the user to use their own model)
We don't collect or share your personal info. We additionally use the GNU Coding Standard, see https://gnu.org/prep/standards
