Code generated in the video can be downloaded from here: github.com/bnsreenu/python_fo... Data set link : www.kaggle.com/datasets/iarun...
Пікірлер: 38
@narayanamurty75864 жыл бұрын
I love your videos. You show me the path when i was in confusion about start off learning deep learning. Keep going , sir. Great job!!!!!
@DigitalSreeni4 жыл бұрын
Glad to hear that
@kartikdube28523 жыл бұрын
This channel is massively underrated! Great work 🔥
@DigitalSreeni3 жыл бұрын
Appreciate it!
@caiyu5382 жыл бұрын
Your series are very helpful. Great videos to learn
@DigitalSreeni2 жыл бұрын
Glad you like them!
@yjoliiyki7063 жыл бұрын
Thanks a lot for your video sir, it's very insightful. I just would like to make sure how many images are in your training datat&test data? Because I m little confused about number 800, during the training, does that mean that your whole dataset has 800 images?
@wong.7772 жыл бұрын
Very good videos. Thank you!
@postcode-x3 жыл бұрын
Hey nice tutorial, thanks. I have a question: suppose I want to output a 2D vector in the geometrical sense (x, y). I've divided the output into three parts: axis, sign, and absolute value. These are the conditions: if one axis value is > 0, then the other axis is assumed to be zero. For example, (0, 1) is a valid vector, (3, 0) is a valid vector, but (3, 4) is not a valid vector. Absolute values are bounded between 0 and 7, for example, (0, 7) is a valid vector, but (-1, 0) does not exist in this case. So, my question is this: should I train the network SEPARATELY for each case, meaning, the first nn will do binary classification of the axis, the second nn will do binary classification of the sign and the third nn will do multiclass classification of the values, with one hot encoded data? Thanks again
@priyankadas71023 жыл бұрын
Really Appreciate and like your videos. Thank You.
@DigitalSreeni3 жыл бұрын
My pleasure!
@leamon9024 Жыл бұрын
Hi, thanks for the tutorial. If possible, could you talk about the implementation of vision transformer(ViT) using keras?
@goopypuff3 жыл бұрын
nice video man, very helpful!
@DigitalSreeni3 жыл бұрын
Glad it helped!
@samarafroz98524 жыл бұрын
Great job sir
@renyuanzhang97064 жыл бұрын
Great content. I have one doubt - your validation_generator only apply rescale of 1/255, meaning it will not increase the data size (unlike train_generator where random transformation is applied). Base on this, 1) does the generator repeats the data in an infinite loop until stopped by epoch? 2) if the answer to (1) is no, can I feed a validation_dataset (instead of a generator) into the model.fit() method (b.t.w the TensorFlow docs claims that "model.fit_generator is deprecated and Please use Model.fit, which supports generators")
@DigitalSreeni4 жыл бұрын
You can use validation data instead of generator, but remember to rescale validation data just like you did for training data. I used generator for validation data to easily mimic operations I performed for training data. I am not aware of TensorFlow depreciating model.fit_generator. I normally use Keras and I've never seen that message. If it is real then you can find alternate library.
@ramidhouib2152 жыл бұрын
Spyser seems to work well with ML , any better IDE?
@junaidlatif28812 жыл бұрын
8:30 no issue. Its part of learning and teaching... ❤
@samranoman6833 жыл бұрын
can you provide me with the dataset please
@minh_tien39113 жыл бұрын
i have an errror, it is in line 39, how to resolve it [WinError 3] The system cannot find the path specified: '/images/cell_images/test/Parasitized/'
@DigitalSreeni3 жыл бұрын
It clearly says that the system cannot find the path. So please make sure the path you gave is correct and also remove the first / before images and try again.
@caiyu5382 жыл бұрын
great lectures
@DigitalSreeni2 жыл бұрын
Glad you like them!
@vineetlokur94952 жыл бұрын
Can I get the dataset?
@erumproductions3 жыл бұрын
Sir on line 39 Syntaxerror :cannot assign to operator Aa rha h mainy apny dataset k two folders bnai hn image_directory = 'C:/Users/OWNER/Desktop/single-grn-dataset/' SIZE = 224 Dataset = [ ] Label = [ ] Infected-Grains-Images = os.listdir(image_directory +'C:/Users/OWNER/Desktop/single-grn-dataset/Infected-Grains/ ') SyntaxError: cannot assign to operator Sir I am new plz guide me
@yahaisha3 жыл бұрын
sir..great video ..tq so much...how to create this binary classification if my data quiet small ? just 200 dataset?
@DigitalSreeni3 жыл бұрын
Try data augmentation but you may run into overfitting. Also try traditional methods such as using Random Forest or SVM, they work better with small datasets. Watch my videos on image classification.