ImageClassifier-Keras

Custom Image Classifier using Custom model and pre-trained models (like MobileNet, VGG, ResNet, EfficientNet,..etc) with help of Tensorflow and Keras.
Its a single tool all you needed for your Image Classification problems.
You can Inference on Test Image, Video and Web-cam using your saved Custom Model or Pre-Trained Model. from a Single Line.
If you need to Save the output Image or Video. Just need to put --save argument in the Last.

Also you can Train Model using Data from S3 Bucket - "Without Downloading"
Its same like Custom and Pre-Trained model Training.
Only few small changes. You need to give extra 3 Arguments. That's all..

Added MLFlow to Model training. So that we can Manage the ML lifecycle,
Including:

Experimentation
Reproducibility
Deployment
Central Model Registry.

Features:

Option to added your own Custom Model
Options to added All Pre-Trained Models avalable in Keras Applications
You can Train your Model using Data from S3 Bucket - "Without Downloading"
MLflow: Managing the end-to-end Machine Learning Lifecycle.

You can Train or Inference your Data in your Custom Model or in any Pre-Trained Model, "in a single line".

Keras Applications

Model	Model Size (MB)	Top-1 Accuracy	Top-5 Accuracy	Parameters	Depth	Time (ms) per inference step (CPU)	Time (ms) per inference step (GPU)
Xception	88	79.0%	94.5%	22.9M	81	109	8.1
VGG16	528	73.1%	90.1%	138.4M	16	69.5	4.2
VGG19	549	71.3%	90.0%	143.7M	19	84.8	4.4
ResNet50	98	74.9%	92.1%	25.6M	107	58.2	4.6
ResNet50V2	98	76.0%	93.0%	25.6M	103	45.6	4.4
ResNet101	171	76.4%	92.8%	44.7M	209	89.6	5.2
ResNet101V2	171	77.2%	93.8%	44.7M	205	72.7	5.4
ResNet152	232	76.6%	93.1%	60.4M	311	127.4	6.5
ResNet152V2	232	78.0%	94.2%	60.4M	307	107.5	6.6
InceptionV3	92	77.9%	93.7%	23.9M	189	42.2	6.9
InceptionResNetV2	215	80.3%	95.3%	55.9M	449	130.2	10.0
MobileNet	16	70.4%	89.5%	4.3M	55	22.6	3.4
MobileNetV2	14	71.3%	90.1%	3.5M	105	25.9	3.8
DenseNet121	33	75.0%	92.3%	8.1M	242	77.1	5.4
DenseNet169	57	76.2%	93.2%	14.3M	338	96.4	6.3
DenseNet201	80	77.3%	93.6%	20.2M	402	127.2	6.7
NASNetMobile	23	74.4%	91.9%	5.3M	389	27.0	6.7
NASNetLarge	343	82.5%	96.0%	88.9M	533	344.5	20.0
EfficientNetB0	29	77.1%	93.3%	5.3M	132	46.0	4.9
EfficientNetB1	31	79.1%	94.4%	7.9M	186	60.2	5.6
EfficientNetB2	36	80.1%	94.9%	9.2M	186	80.8	6.5
EfficientNetB3	48	81.6%	95.7%	12.3M	210	140.0	8.8
EfficientNetB4	75	82.9%	96.4%	19.5M	258	308.3	15.1
EfficientNetB5	118	83.6%	96.7%	30.6M	312	579.2	25.3
EfficientNetB6	166	84.0%	96.8%	43.3M	360	958.1	40.4
EfficientNetB7	256	84.3%	97.0%	66.7M	438	1578.9	61.6
EfficientNetV2B0	29	78.7%	94.3%	7.2M	-	-	-
EfficientNetV2B1	34	79.8%	95.0%	8.2M	-	-	-
EfficientNetV2B2	42	80.5%	95.1%	10.2M	-	-	-
EfficientNetV2B3	59	82.0%	95.8%	14.5M	-	-	-
EfficientNetV2S	88	83.9%	96.7%	21.6M	-	-	-
EfficientNetV2M	220	85.3%	97.4%	54.4M	-	-	-
EfficientNetV2L	479	85.7%	97.5%	119.0M	-	-	-

So, Let's Get Started...

Clone this Repository

git clone https://github.com/naseemap47/ImageClassifier-Keras.git

Install Depencies

cd ImageClassifier-Keras
pip3 install -r requirements.txt

Collect Data

Inside Data Directory. The name of Class Directory should be in index number (starting from zero). Inside each class having its corresponding Image Data.

Edit classes.txt file

Put your Class names in the order corresponding to index number names given to Class Directories.
Example: classes.txt

cat
dog
daisy
dandelion
roses
sunflowers
tulips
.....

Example: Dataset Structure:

├── Dataset
│   ├── 0
│   │   ├── 1.jpg
│   │   ├── 2.jpg
│   │   ├── ...
│   ├── 1
│   │   ├── 1.jpg
│   │   ├── 2.jpg
│   │   ├── ...
.   .
.   .

Model Training

You have 3 Options:

Train on your own Custom Model
Train on Pre-Trained Model
Train Model using Data from S3 Bucket - "Without Downloading"

1. Train on your own Custom Model

Open Models.py:
Edit custom_model function to your own Custom Model Function.

-i, --dataset: Path to dataset/dir
-s, --img_size: Size of Image used to train the model
-b, --batch_size: Batch Size of Model Training
-e, --epochs: Epochs of Model Training
--model: Select Model type custom or mobilenetV2,..etc

Custom Model: custom

--model_save: Path to save model.h5

Example:

python3 train.py --dataset Data/ --img_size 32 --batch_size 16 \
                 --epochs 50 --model custom --model_save model.h5

Train Model using Data from S3 Bucket - "Without Downloading"

Its same like Custom model Training. Only few small changes. You need to give extra 3 Arguments. That's it..

-n, --bucket: S3 Bucket Name
-i, --dataset: path to dataset/dir in S3 Bucket
-s, --img_size: Size of Image used to train the model
-b, --batch_size: Batch Size of Model Training
-e, --epochs: Epochs of Model Training
--model: Select Model type custom or mobilenetV2,..etc

Custom Model: custom

--model_save: Path to save model.h5
--aws_region: AWS Region Name
--aws_access_key_id: AWS Access Key ID
--aws_sec_access_key: AWS Secret Access Key

Example:

python3 trainS3.py --bucket my_bucket --dataset ImageData/Train --img_size 320 \
                   --batch_size 8 --epochs 100 --model custom \
                   --model_save model.h5 --aws_region 'ap-south-1' \
                   --aws_access_key_id 'AKDU74NH7MNO2NFEY5NK' \
                   --aws_sec_access_key 'JtyTgFFr/34huiHygUGu&hg7UIHisdhsoit7dsHF'

2. Train on Pre-Trained Model

🗒️ Note:

I added a Dense Layer on last Layer in the Pre-Trained Model, with size 1024. You can edit that layer, If necessary.
To Edit:
Open Models.py, go to Line-192 of pre_trainied_model function, there you can find the Dense Layer.

-i, --dataset: Path to dataset/dir
-b, --batch_size: Batch Size of Model Training
-e, --epochs: Epochs of Model Training
--model: Select Model type custom or mobilenetV2,..etc

Xception: xception
VGG16: vgg16
VGG19: vgg19
ResNet50: resnet50
ResNet101: resnet101
ResNet152: resnet152
ResNet50V2: resnet50V2
ResNet101V2: resnet101V2
ResNet152V2: resnet152V2
InceptionV3: inceptionV3
InceptionResNetV2: inceptionresnetV2
MobileNet: mobilenet
MobileNetV2: mobilenetV2
MobileNetV3Small: mobilenetV3Small
MobileNetV3Large: mobilenetV3Large
DenseNet121: densenet121
DenseNet169: densenet169
DenseNet201: densenet201
NASNetLarge: nasnetLarge
NASNetMobile: nasnetMobile
EfficientNetB0: efficientnetB0
EfficientNetB1: efficientnetB1
EfficientNetB2: efficientnetB2
EfficientNetB3: efficientnetB3
EfficientNetB4: efficientnetB4
EfficientNetB5: efficientnetB5
EfficientNetB6: efficientnetB6
EfficientNetB7: efficientnetB7
EfficientNetV2B0: efficientnetV2B0
EfficientNetV2B1: efficientnetV2B1
EfficientNetV2B2: efficientnetV2B2
EfficientNetV2B3: efficientnetV2B3
EfficientNetV2S: efficientnetV2S
EfficientNetV2M: efficientnetV2M
EfficientNetV2L: efficientnetV2L

--model_save: Path to save model.h5

Image Size set automatically with respect to selected Pre-Trained Model

Example:

python3 train.py --dataset Data/ --batch_size 8 --epochs 80 \
                 --model efficientnetB1 --model_save model.h5

Train Model using Data from S3 Bucket - "Without Downloading"

Its same like Pre-Trained model Training. Only few small changes. You need to give extra 3 Arguments. That's it..

-n, --bucket: S3 Bucket Name
-i, --dataset: path to dataset/dir in S3 Bucket
-b, --batch_size: Batch Size of Model Training
-e, --epochs: Epochs of Model Training
--model: Select Model type custom or mobilenetV2,..etc

All types of pre-trained models given above

--model_save: Path to save model.h5
--aws_region: AWS Region Name
--aws_access_key_id: AWS Access Key ID
--aws_sec_access_key: AWS Secret Access Key

Example:

python3 trainS3.py --bucket my_bucket --dataset ImageData/Train \
                   --batch_size 8 --epochs 100 --model vgg16 \
                   --model_save model.h5 --aws_region 'ap-south-1' \
                   --aws_access_key_id 'AKDU74NH7MNO2NFEY5NK' \
                   --aws_sec_access_key 'JtyTgFFr/34huiHygUGu&hg7UIHisdhsoit7dsHF'

MLFlow UI

MLflow is an open source platform for managing the end-to-end machine learning lifecycle
terminal is in the same directory that contains mlruns, and type the following:

mlflow ui

# OR
mlflow ui -p 1234

The command mlflow ui hosts the MLFlow UI locally on the default port of 5000.
However, the options -p 1234 tell it that you want to host it specifically on the port 1234.

open a browser and type in http://localhost:1234 or http://127.0.0.1:1234

Inference

🗒️ Note:
Please make sure that, You Trained and Inferencing in the same Python Version.
If NOT, there is a chance to get Error: "code = marshal.loads(raw_code) ValueError: bad marshal data (unknown type code)".
I got an Error when loading "InceptionResNetV2" trained model. Becouse I trained on Colab (python 3.7.14) and Inference in (python 3.9).
The rest of the model didn't get any error. Only I got in "InceptionResNetV2" Model.

Explanation:
This typically happens when you save a model in one Python version (e.g., 3.6) and then try to load that model in another Python version (e.g., 3.9), as the binary serialization that Keras uses (marshal) is not upwards/downwards compatible. Try to install an old version of Python with an appropriate version of the Tensorflow / Keras libraries. If the model was not trained by yourself, you may ask the creators to export the trained models in a different format that doesn't have these problems, like ONNX.

So, its better to train and Inference in the same Python Version.

You have 2 Options:

Inference on your own Custom Model
Inference on Pre-Trained Model

1. Inference on your own Custom Model

--img_size: Size of Image used to Train the model
-m, --model: Path to saved .h5 model, eg: dir/model.h5
--model_type: Select model type custom or mobilenetV2,vgg16..etc
-c, --conf: Min prediction confidence to detect pose class (0<conf<1)
--source: Path to sample image, video or Webcam
--save: If need to Save image, video or Webcam

Example:

Image

python3 inference.py --img_size 32 --model model.h5 --model_type custom \
                     --conf 0.8 --source test/image.jpg

# to Save output
python3 inference.py --img_size 32 --model model.h5 --model_type custom \
                     --conf 0.8 --source test/image.jpg --save

Video

python3 inference.py --img_size 32 --model model.h5 --model_type custom \
                     --conf 0.8 --source test/video.mp4

# to Save output
python3 inference.py --img_size 32 --model model.h5 --model_type custom \
                     --conf 0.8 --source test/video.mp4 --save

Web-cam

python3 inference.py --img_size 32 --model model.h5 --model_type custom \
                     --conf 0.8 --source 0

# to Save output
python3 inference.py --img_size 32 --model model.h5 --model_type custom \
                     --conf 0.8 --source 0 --save

2. Inference on Pre-Trained Model

-m, --model: Path to saved .h5 model, eg: dir/model.h5
--model_type: Select model type custom or mobilenetV2,vgg16..etc

Xception: xception
VGG16: vgg16
VGG19: vgg19
ResNet50: resnet50
ResNet101: resnet101
ResNet152: resnet152
ResNet50V2: resnet50V2
ResNet101V2: resnet101V2
ResNet152V2: resnet152V2
InceptionV3: inceptionV3
InceptionResNetV2: inceptionresnetV2
MobileNet: mobilenet
MobileNetV2: mobilenetV2
MobileNetV3Small: mobilenetV3Small
MobileNetV3Large: mobilenetV3Large
DenseNet121: densenet121
DenseNet169: densenet169
DenseNet201: densenet201
NASNetLarge: nasnetLarge
NASNetMobile: nasnetMobile
EfficientNetB0: efficientnetB0
EfficientNetB1: efficientnetB1
EfficientNetB2: efficientnetB2
EfficientNetB3: efficientnetB3
EfficientNetB4: efficientnetB4
EfficientNetB5: efficientnetB5
EfficientNetB6: efficientnetB6
EfficientNetB7: efficientnetB7
EfficientNetV2B0: efficientnetV2B0
EfficientNetV2B1: efficientnetV2B1
EfficientNetV2B2: efficientnetV2B2
EfficientNetV2B3: efficientnetV2B3
EfficientNetV2S: efficientnetV2S
EfficientNetV2M: efficientnetV2M
EfficientNetV2L: efficientnetV2L

-c, --conf: Min prediction confidence to detect pose class (0<conf<1)
--source: Path to sample image, video or Webcam
--save: If need to Save image, video or Webcam

Image Size set automatically with respect to selected Pre-Trained Model

Example:

Image

python3 inference.py --model model.h5 --model_type mobilenet \
                     --conf 0.8 --source test/image.jpg

# to Save output
python3 inference.py --model model.h5 --model_type mobilenet \
                     --conf 0.8 --source test/image.jpg --save

Video

python3 inference.py --model model.h5 --model_type mobilenet \
                     --conf 0.8 --source test/video.mp4

# to Save output
python3 inference.py --model model.h5 --model_type mobilenet \
                     --conf 0.8 --source test/video.mp4 --save

Web-cam

python3 inference.py --model model.h5 --model_type mobilenet \
                     --conf 0.8 --source 0

# to Save output
python3 inference.py --model model.h5 --model_type mobilenet \
                     --conf 0.8 --source 0 --save