- Видео 202
- Просмотров 395 821
Eran Feit
Израиль
Добавлен 26 мар 2013
See the world through data and AI
Welcome to my channel, where I explore and publish cool Python projects.
You can follow the world of Computer Vision, TensorFlow, Keras, and Python with my tutorials.
I am focused on convolutional neural networks , and pre-trained models for object classifications object detections, and segmentations as well.
You are most welcome to contact me and share the videos :
~~~~~~~~~~~~~~~ CONNECT ~~~~~~~~~~~~~~~
☕ Buy me a coffee - ko-fi.com/eranfeit
🖥️ Email : feitgemel@gmail.com
🌐 eranfeit.net
🤝 Fiverr : www.fiverr.com/s/mB3Pbb
🐦 Twitter - eran_feit
📸 Instagram - eran_feit
▶️ Subscribe - youtube.com/@eranfeit?sub_confirmation=1
🐙 Facebook - groups/3080601358933585
📝 Medium - medium.com/@feitgemel
Eran Feit
#Eranfeit #Python #convolutionalneuralnetworks #tensorflow #transferlearning
Welcome to my channel, where I explore and publish cool Python projects.
You can follow the world of Computer Vision, TensorFlow, Keras, and Python with my tutorials.
I am focused on convolutional neural networks , and pre-trained models for object classifications object detections, and segmentations as well.
You are most welcome to contact me and share the videos :
~~~~~~~~~~~~~~~ CONNECT ~~~~~~~~~~~~~~~
☕ Buy me a coffee - ko-fi.com/eranfeit
🖥️ Email : feitgemel@gmail.com
🌐 eranfeit.net
🤝 Fiverr : www.fiverr.com/s/mB3Pbb
🐦 Twitter - eran_feit
📸 Instagram - eran_feit
▶️ Subscribe - youtube.com/@eranfeit?sub_confirmation=1
🐙 Facebook - groups/3080601358933585
📝 Medium - medium.com/@feitgemel
Eran Feit
#Eranfeit #Python #convolutionalneuralnetworks #tensorflow #transferlearning
How to make a Real-time Deep Fake | Face Swap Tutorial
Let's learn how to easily make a real-time deep fake with the amazing software Deep Face Live.
You can swap your face from a webcam or the face in the video using trained face models and a single image
What you’ll learn:
* How to install the environment and the required Python libraries
* how to make deep fake with 3 alternative
You can find the instructions file here : ko-fi.com/s/20137b6a74
More computer vision tutorials in my blog page : eranfeit.net/blog/
You can find more projects and tutorials in this playlist : ruclips.net/p/PLdkryDe59y4ZW-E59KQYegP5dr1WYnbpn
~~~~~~~~~~~~~~~ recommended courses and books ~~~~~~~~~~~~~~~
A perfect course for learning modern Computer Vision with deep dive i...
You can swap your face from a webcam or the face in the video using trained face models and a single image
What you’ll learn:
* How to install the environment and the required Python libraries
* how to make deep fake with 3 alternative
You can find the instructions file here : ko-fi.com/s/20137b6a74
More computer vision tutorials in my blog page : eranfeit.net/blog/
You can find more projects and tutorials in this playlist : ruclips.net/p/PLdkryDe59y4ZW-E59KQYegP5dr1WYnbpn
~~~~~~~~~~~~~~~ recommended courses and books ~~~~~~~~~~~~~~~
A perfect course for learning modern Computer Vision with deep dive i...
Просмотров: 829
Видео
What if there was a tool that could enhance old Video to a FULL HD ?
Просмотров 722 месяца назад
Do you have old videos that you wish looked better? In this video I'm going to show you how to restore the quality of your old videos Video Enhancer free AI tool This process is really simple and easy, But you need a high GPU card , and patience for the transformation. This tutorial based on DiffIR2VR-Zero: Zero-Shot Video Restoration with Diffusion-based Image Restoration Models What you’ll le...
This FREE Image to Video Ai Can Control Expressions and Emotion | Live Portrait Tutorial
Просмотров 932 месяца назад
In this video I will share ai video generator that can animate your still images with realistic expressions and emotions. This is called live portrait. This free image to video ai is game changer. What you’ll learn: * How to install the environment and the required Python libraries * How to Generate video animations based on your images and driving face videos You can find the instructions and ...
Create AI Images with Your Face ! (Free and Easy)
Просмотров 1722 месяца назад
In this tutorial, We will Learn how to use Photomaker to generate stunning photos based on your own face. We will cover the installation process and how to use your own images with text prompt to generate new cool images . What you’ll learn: * How to install the environment and the required Python libraries * How to Generate photos of your own face You can find the instructions and the demo fil...
How I Combined 3 Powerful AI Models to Automate Image Editing
Просмотров 1242 месяца назад
In this tutorial, we explore AI-driven image processing using Grounding DINO, SAM (Segment Anything Model), Stable Diffusion and your own prompt !!!! This step-by-step guide covers everything from model loading and image preprocessing to detecting, segmenting, and generating new images. What you’ll learn: * How to install the environment and the required Python libraries * How to use GroundingD...
Free Subtitle Generator - The FASTEST way to create Subtitles
Просмотров 2193 месяца назад
Faster-whisper is a reimplementation of OpenAI's Whisper model using CTranslate2, which is a fast inference engine for Transformer models. This implementation is up to 4 times faster than openai/whisper for the same accuracy while using less memory. We will learn how to implement automatic transcription of audio and video, allowing us to generate SRT subtitles. Additionally, we will learn how t...
Remove Objects from Images Like Magic - Step-by-Step Guide
Просмотров 1203 месяца назад
Automated object remover Inpainter is a project that combines Semantic segmentation and EdgeConnect architectures with minor changes in order to remove specified objects from photos. For Semantic Segmentation, the code from Pytorch has been adapted, whereas for EdgeConnect, the code has been adapted from This project is capable of removing objects from list of 20 different ones. It can be used ...
FaceFusion Face Swap Is WILD (Full FaceFusion Installation and Tutorial)
Просмотров 1,1 тыс.3 месяца назад
Face Swapping in Images & Videos Using FaceFusion | Full Setup & Tutorial In this tutorial, we dive into the world of face swapping, using the powerful FaceFusion library. You'll see hands-on examples of how to swap faces in both images and videos, followed by a step-by-step guide to setting up a fresh environment What you’ll learn: * How to install the environment and the required Python libra...
Making AI Generated Image Captions using Python | Automatic Image Captioning with Vit-Gpt2
Просмотров 1043 месяца назад
In this tutorial, we will use Hugging Face’s pre-trained 'nlpconnect/vit-gpt2-image-captioning' model. You'll learn how to easily generate descriptive captions for any image using Python and PyTorch. We’ll walk you through setting up the Vision Transformer (ViT) for image processing and GPT-2 for text generation. What you’ll learn: * How to install the environment and the required Python librar...
Object Detection Heatmap using YoloV8 | Dogs detection
Просмотров 1564 месяца назад
An object detection heatmap is a visual representation used to highlight areas within an image where an object detection model has identified the presence of objects In this video, we will generate heatmaps using object detection and object tracking . The tutorial is based on YoloV8 Code for the tutorial : ko-fi.com/s/21cb5fdeda You can find more computer vision tutorials in my blog page : eran...
How to Detect Dogs using Deep learning
Просмотров 1484 месяца назад
In this tutorial, we delve into the world of computer vision by creating a YOLOv8-based dogs breed object detection using Deep learning . What you'll learn: How to import and utilize the YOLOv8 model from the Ultralytics library. How to convert annotations from XML format to Yolo format How to load and interpret annotation data from a YAML file. How to train your YOLOv8 model with a dogs datase...
How to Detect Basketball Game using Deep Learning
Просмотров 4414 месяца назад
In this tutorial, we delve into the world of computer vision by creating automatic labeling in Video streaming , and YOLOv8-based basketball object detection using Deep learning . We'll work with game footage from a RUclips match between Real Madrid and Maccabi Tel-Aviv to build a model capable of detecting players, referees , and the basketball itself in real-time. This tutorial covers the ful...
No Dataset? No Problem! Create a Horse Race Detection Model using Deep learning
Просмотров 3984 месяца назад
in this video, We will learn how to train a horse race classes , even though we do not have a dataset. . The pipeline includes video frame extraction, object detection model training, and displaying the results with bounding boxes and labels. Here’s an overview of what each part of the code does: 1. Frame Extraction from Video : The first part of the code extracts frames from a horse race video...
How to Detect Under the Sea classes using Deep learning | Yolo-Nas
Просмотров 835 месяцев назад
YOLO-NAS delivers state-of-the-art (SOTA) performance with the unparalleled accuracy-speed performance, outperforming other models such as YOLOv5, YOLOv6, YOLOv7 and YOLOv8. Training with SuperGradients, PyTorch-based computer vision library. SuperGradients is fully compatible with PyTorch Datasets and Dataloaders, so you can use your dataloaders as is. In this video, we demonstrate how to impl...
Yolo-Nas Step by Step Guide to Object Detection
Просмотров 2425 месяцев назад
YOLO-NAS is a game-changer in object detection model developed by Deci and delivers superior real-time object detection capabilities and production-ready for production. The primary claim of Yolo-NAS is that it can detect small objects better than the previous models. In this computer vision tutorial video, we will run inference of Yolo-NAS model on images. We will learn two methods for extract...
How to Significantly Enhance Yolov8 with SAHI
Просмотров 6025 месяцев назад
How to Significantly Enhance Yolov8 with SAHI
How to Detect Bone Fracture using Deep Learning
Просмотров 2,2 тыс.5 месяцев назад
How to Detect Bone Fracture using Deep Learning
How to Build a Deep learning model to detect Ships
Просмотров 3025 месяцев назад
How to Build a Deep learning model to detect Ships
How to Detect Cards using Deep Learning | Yolo V8
Просмотров 8305 месяцев назад
How to Detect Cards using Deep Learning | Yolo V8
How to Detect Objects from Youtube Video in Real Time
Просмотров 3125 месяцев назад
How to Detect Objects from RUclips Video in Real Time
Teeth Detection using Deep learning | Yolo V8
Просмотров 5556 месяцев назад
Teeth Detection using Deep learning | Yolo V8
How to Train YOLOv5 on a Custom Dataset
Просмотров 3586 месяцев назад
How to Train YOLOv5 on a Custom Dataset
Object Detection in 15 minutes with YOLOv5 & Python !
Просмотров 5106 месяцев назад
Object Detection in 15 minutes with YOLOv5 & Python !
YOLOX Object Detection without coding
Просмотров 2376 месяцев назад
YOLOX Object Detection without coding
Object detection Using Detection Transformer (Detr) for Bone fraction dataset
Просмотров 5626 месяцев назад
Object detection Using Detection Transformer (Detr) for Bone fraction dataset
Train Your Own Object Detector with Detectron2
Просмотров 1407 месяцев назад
Train Your Own Object Detector with Detectron2
Exploring Detectron2 For easy Object Detection
Просмотров 1607 месяцев назад
Exploring Detectron2 For easy Object Detection
Build an Object Detector using SSD MobileNet v3
Просмотров 1,4 тыс.7 месяцев назад
Build an Object Detector using SSD MobileNet v3
Say Goodbye to Manual Labeling: YOLOv8 Auto-Label Segmentation Made Easy
Просмотров 3917 месяцев назад
Say Goodbye to Manual Labeling: YOLOv8 Auto-Label Segmentation Made Easy
Can use custom image to train the model?
Yes Look at this playlist. You can choose many model types. Include Detectron2 ruclips.net/p/PLdkryDe59y4bXa-1wOEAF4KljIMamhWd0&si=LGrnSMKlJ3IFYV8f
I am trying to use this to have a photo on a dam or off a dam in a floodplain, detect water level increases and report back to the office through notifications
Ok. And did you succeed ? You can send me your dataset to my email and I would look into it.
No, I’m still trying to figure out if anyone has done this so I don’t need to create any software from scratch
@@russm195 I can try code the model if you send me the dataset and samples of the required result.
Dear Eran , if we wanna detect the waste vs non waste images how can we train the model by using mobilenetv2
Hi. Please define waste images.
is detr better than yolo
I am making a new video about Detr , as for accuracy it’s seems the same. As for time of training , it seems need more time of training.
I am using the ResNet50 Model for my image classifyer and i always get the problem, that the leftover batch which is smaller than the specified batch_size is causing problems. Is there another way to handle this instead of dropping the final batch? I can't find anything online
Never met this error message.
Nice video thanks you
Thanks. You are welcome to subscribe :)
Sir I'm stuck in 15:00.i'm getting error when applied "from keras import utils" also. What is the import statement??? ... y_train=????(y_train,num_classes=7)
Hi Send me your full code to my email , I will look into it
How can I run this on a live video feed?
It is possible. Replace the image with collecting frames from a camera.
can efficientnet v2 do 2 class binary classification?
Yes, You can adjust the code : # If binary classification, replace `softmax` with `sigmoid`: model = tf.keras.Sequential([ base_model, layers.GlobalAveragePooling2D(), layers.Dense(1024, activation='relu'), layers.Dropout(0.5), layers.Dense(1, activation='sigmoid') # use 1 output node for binary classification ]) # Update the loss to binary_crossentropy for binary classification: model.compile(optimizer=adam_opt, loss='binary_crossentropy', metrics=['accuracy']) This update will output a single probability score for each image (interpreted as the probability of belonging to one of the two classes) Good luck
@@eranfeit tysm its work
How do I change how many layers I want to freeze?
Modify the code: # Define how many layers you want to freeze num_layers_to_freeze = 100 # Adjust this number as needed # Freeze the specified number of layers for layer in model.layers[:num_layers_to_freeze]: layer.trainable = False -> num_layers_to_freeze controls how many layers you freeze, starting from the first layer up to the specified count. Adjust num_layers_to_freeze based on your requirements. For example, setting it to 100 will freeze the first 100 layers, while the rest will remain trainable.
Wait i cannot find the wav2lip.gan.pth file download link in the description Could you please help me out with it
Exactly what I was looking for! Thanks so much for sharing this.
Thanks
thanks for the video, an ingestible duration, no fluff, the music, it's all good with one small note - the font size is too small, ctrl-shift-+ would help a lot.
Thanks. I will remember that in my next videos.
רק לפי המבטא זיהיתי וגם לפי השם :) בכל אופן אם תרצה יש לי סקריפט שעושה עבודה דיי טובה ביצירת כתוביות לפי פרמטרים מעודכנים. התוצאות כמעט מושלמות
הי בדרך כלל אני כותב ומאמן מודלים בעצמי. בסרטונים האחרונים בדקתי ושחקתי עם כמה ספריות מוכנות. תודה
Can it be used for youtube live streaming?
Yes. You should replace the source to your usb camera
@@eranfeit Thank you, Sir. Can it be used on trained models and other videos?
@@EkaSantiFadhilah Your question is to general. You can find more tutorials for Object detection training here : ruclips.net/video/e-tfaEK9sFs/видео.html
Hi, how can i use ssd mobile net with my own data??, your advice would be huge help man.
Hi, Never tried with SSD . I believe it is a CNN architecture , so you have to "play" with the layers. Do some google search . I can see that someone asked it before : stackoverflow.com/questions/43493806/how-do-i-retrain-ssd-object-detection-model-for-our-own-dataset
my pc is amd .I havent cuda .How can I do ?
Hi, Pytorch supports GPU cards with Cuda. Moreover , It also support CPU card . Try install Pytorch with this command instead : # CPU Only conda install pytorch==2.0.1 torchvision==0.15.2 torchaudio==2.0.2 cpuonly -c pytorch Eran
niceee
Thanks, You are welcome to subscribe ..
MiDas/model.pt doesnt exists, can u help me ?
I did not understand your question . Which link does not work. The "midas\model.pt" is the name of the model after the move command which move and rename the file .
Thanks Mate, its worked thats exactly what i need !
Thanks
amazing
Thanks
nice
Thanks
teşekkürler Eran, değiştirmem gereken videoda 2 adet insan var. 2 adet resim var. tek seferde yapamıyorum. ipucu verirmisin?
No. Just run it in two sessions , while the output of session one will be the input of the second
Hello, I want to use it for custom data. But I can't find a source. Can you guide me? Thanks a lot:)
Detection is a little bit “oldy” Try this one : ruclips.net/video/JbEy4Eefy0Y/видео.htmlsi=5LRcNbhWx2sSevd4
But why gpu doesn't work on Windows ? My windows laptop has GTX 1050 Ti, why cant we use gpu for detectron2 ?
You are right. It is a problem in the Detectron2 . I did not manage to run in using the GPU Pytorch version. You can always install WSL2 in the Windows environment , to run a Linux on your machine
also, is it possible to use anaconda environment for this?
Yes, Watch the installation part (from 00:48) . You can see it was installed using Anaconda environment
would this work on small images like 224x224? and woul it give good results if i have not the best quality of photos?
By default, the images will be saved at resolution of 1024x1024, the original output size of StyleGAN. If you wish to save outputs resized to resolutions of 256x256, you can do so by adding the flag --resize_outputs.
Do i need cuda to follow this tutorial?
Cuda is better for performance. You can always install the PyTorch version that supports CPU ( and not Gpu )
@@eranfeit oh thank you for eloborating. I was planning to use your tutorial for my very low league basketball club, to record our games and then automatically create highlights for it. I saw that the two 5-minute highlights already created frame images of 8,5GB. So ill probably run into Size and time issues when i record a whole game. But thanks for the tutorial, very insightful
@@GSTamer Thanks and good luck
Awesome video ,Thanks
Thanks. You are welcome to subscribe :)
ValueError: Data cardinality is ambiguous. Make sure all arrays contain the same number of samples.'x' sizes: 124 'y' sizes: 14 I am getting ths error please resolve
Did you download the code ?
No
where to download??
Github code??
@@sayalisalkade9248 Code for this video: ko-fi.com/s/3f1fdb4316
dataset link?
Hi Here is the link : www.kaggle.com/datasets/slavkoprytula/aquarium-data-cots
What happen if I don't wanna use a wieght for first instance?
Can you elaborate your question ?
Very good
Thank you :)
wheres the repo folder?
You can find a link for the instructions for this tutorial in the video description.
Apreciate Sir. It was a great and easy description. 👍👍👍
Thanks. You are welcome to subscribe.
How can you get test accuration? Can you give me that code?
There is a link for the code in the video description
@@eranfeit But there is no code to calculate the test accuration. If you know, can you give it?
I need it for undergraduate thesis purpose
Can't we just copy it from the video?
Amazing tutorial! Thanks a lot :)
Thanks. You are welcome to subscribe my newsletter in the bottom of this link :) eranfeit.net
Will you do decayed detecion?
I did not understand your question
😗😗🤔nice👍👍👍👍👍
Thanks :)
where can we get the dataset can you link it in description
Hi, I updated the video description with the link to the dataset
@@eranfeit thanks
@@uncover_ai You care welcome to subscribe , and share it among your friends
Can you do heatmap for football match ?
I believe the answer is yes. Did you try it on a football video ?
Actually no @@eranfeit
how can i implement detection for example when the initial cat sound started and ended how can i get this information in the end result ( like in result it should show cat sound - 2.3s)
This is a pre trained model from Tensor hub. I believe the model was trained based on a permanent audio length , so the inference looks for audio in similar length Try to read the documentation if it returns the time value for detection. You can try to build your own audio detection.
Greetings I’m working on my final year project called ShieldVision: Advanced Theft Detection & Alerting System. It uses AI for object detection, real-time video capture, and SMS alerts to enhance theft prevention. I need your guidance on a few things: What should I focus on to learn image processing and object detection, especially for gun detection? Can you recommend any good courses or resources on AI in security systems? What are the best practices for implementing an object detection system from scratch? What hardware would you recommend, considering I’m on a budget? Your advice would be a huge help! Thank you in advance!
Hi. You need learn : How to capture frames from a camera ? Build a model for detection objects that indicates theft. You need a dataset for that. If you have the dataset I can help with that. Yolo can do the job Then you need to code a prediction inference for your model That would detect the on fresh images or using a camera. I can help you with that. And at last you should code sms messaging. Eran
Can you do Car crush detection algorithm with yolo?
I think it less interesting. Do you have a dataset ?
@@eranfeit There is a dataset in roboflow, but it is not a very nice dataset. I tried it, but even if there was no accident, the algorithm perceived it as an accident. It is a bit of a difficult problem. I guessed by calculating the central point distance and angles of the cars, there were some problems here without a dataset, since the cars will be side by side in traffic, the model will guess wrong at the red light.
Great Video! Looking forward to a playlist on Intro To Computer vision. I hope it happens someday😄
Thanks, I am more focused on Use cases and models. Maybe , I will take a short break for demonstrate cool Python computer vision pre-trained libraries , found it Git-hub
watching at 1.5 speed
If you mange to follow , great.
Can we do automatic labeling for objects in different data sets?
Yes
Good job Eran
Thank you
You are the best, thank you
Thanks You are to subscribe to my monthly newsletter at this page : eranfeit.net ( go to the bottom of the first page ) Eran