TFLite Object Detection Android App Tutorial Using YOLOv5 | Yolov5 to tflite conversion

Learn to Develop your Object Detection Mobile Application from yolov5 model.
Github:

Topics covered:
1- Train yolov5 model
2- Convert yolov5 (.pt model) into a tensorflow model(.pb file)
3- Convert tensorflow model (.pb model) to tflite model.
4- Download and install Android Studio
5- Build and run your Object detection App.

YOLOv5 belongs to the YOLO family of models. YOLO was initially introduced as the first object detection model that combined bounding box prediction and object classification into a single end to end differentiable network. It was written and is maintained in a framework called Darknet. YOLOv5 is the first of the YOLO models to be written in the PyTorch framework and it is much more lightweight and easy to use. That said, YOLOv5 did not make major architectural changes to the network in YOLOv4 and does not outperform YOLOv4 on a common benchmark, the COCO dataset.

[pavanchandravundurthy]

Thank you so much! The content was crisp and highly beneficial.

[benyajamiu]

Great Job, even though i'm sleepy but still watched the YouTube...Well done...

[ashberten]

Thanks for the tutorial. But the detection performance is when you test it using the tflite model. Could you suggest the reason behind this, and how to improve it?

[samandarkhasanov]

I made a custom model. It identifies 3 situations. I placed model and label.txt in android studio assets section. Even if I write 3 labels in the txt file, it detects only 2. What can this be and how can it be solved? Please help me

[vglzuoz]

Could you please share your bulid gradle configurations and the jdk you used for the app?It would mean a lot to me.

[ismailhamouda]

this code dose not run on yoloV8 because output sahpe in yolo8 is defferent, اow is this problem solved?

[aryaafshar]

thanks for your video there is issue on Java Version

[ercumentcozer]

no variants found for ':app'. i am getting the error. How should I set the versions?

[haticeozdemir]

Hello, I want to test on a model I trained myself, but the application crashes? What is the reason?

[ayushgupta-bodu]

Can you explain which code enable the camera to detect object ?
Please reply

[cuak-cuack]

why the predictions of your objects in runs/detect/exp9/ are not related with the categories of your custom dataset but instead with the coco dataset categories (like cars, stop_sign, etc)?

[yashh]

why tflite model's results are wrong?

[amiralizadeh]

Thank you for your excellent training. I have a problem at the last stage. When I run and open the program, it suddenly stops and the program closes.

[naziafarhat-myvl]

same topic covered in this video for yolov9 is highly requested.

[wdykpty]

Thanks for informative vid, could i ask question? The result of inference on android phone could be use for control of motor like aduino?

[PiyushRaj-onen]

Hii...i successfully converted yolov5 to tflite..but when i want infrencing of tflite model, its output_tensor shape is (1, 25600, 17)...so how can i fix this issue...??

[pinocchio]

Hello! This is now outdated and i can't run properly in android. Do you have a latest version of this? :))

[Andi]

How do I add that when an object has been detected, the detection results will be saved in history🙏

[yhendra]

i use the same code and anything but, why my program can't detect the image after training test?

[syalwadea]

thank you for your video, if i want convert model yolo to tflite but not start from training (i have already model), why the results of detection didn't same like result from .pt model? :(