How to increase accuracy of model from tensorflow model zoo? Unicorn Meta Zoo #1: Why another podcast? Announcing the arrival of Valued Associate #679: Cesar Manara 2019 Moderator Election Q&A - Questionnaire 2019 Community Moderator Election ResultsTensorflow oscillating Test and Train Accuracy?Caffe net.predict() , predict same probabilityConvnet training error does not decreaseDifficulty in choosing Hyperparameters for my CNNConvNet exploding/vanishing lossSSD based on ResNet-101 doesn't improve over SSD-VGGNetWhy is my Keras model not learning image segmentation?How is Stochastic Gradient Descent done in Faster RCNN?Odd Loss Curves for Object Detection TaskTraining deep CNN with noisy dataset
All ASCII characters with a given bit count
How long after the last departure shall the airport stay open for an emergency return?
"Whatever a Russian does, they end up making the Kalashnikov gun"? Are there any similar proverbs in English?
Error: Syntax error. Missing ')' for CASE Statement
Are all CP/M-80 implementations binary compatible?
Split coins into combinations of different denominations
Putting Ant-Man on house arrest
What is it called when you ride around on your front wheel?
Why didn't the Space Shuttle bounce back into space as many times as possible so as to lose a lot of kinetic energy up there?
What's parked in Mil Moscow helicopter plant?
What was Apollo 13's "Little Jolt" after MECO?
What if Force was not Mass times Acceleration?
Why did C use the -> operator instead of reusing the . operator?
Trumpet valves, lengths, and pitch
Raising a bilingual kid. When should we introduce the majority language?
Co-worker works way more than he should
Do you need a weapon for Thunderous Smite, and the other 'Smite' spells?
Is accepting an invalid credit card number a security issue?
Is it acceptable to use working hours to read general interest books?
Passing args from the bash script to the function in the script
Does Feeblemind produce an ongoing magical effect that can be dispelled?
Can I criticise the more senior developers around me for not writing clean code?
Align column where each cell has two decimals with siunitx
Why did Israel vote against lifting the American embargo on Cuba?
How to increase accuracy of model from tensorflow model zoo?
Unicorn Meta Zoo #1: Why another podcast?
Announcing the arrival of Valued Associate #679: Cesar Manara
2019 Moderator Election Q&A - Questionnaire
2019 Community Moderator Election ResultsTensorflow oscillating Test and Train Accuracy?Caffe net.predict() , predict same probabilityConvnet training error does not decreaseDifficulty in choosing Hyperparameters for my CNNConvNet exploding/vanishing lossSSD based on ResNet-101 doesn't improve over SSD-VGGNetWhy is my Keras model not learning image segmentation?How is Stochastic Gradient Descent done in Faster RCNN?Odd Loss Curves for Object Detection TaskTraining deep CNN with noisy dataset
$begingroup$
Situation:
My dataset is 70k images of people wearing clothes. Images are labeled: bbox position and class. There are 10 classes. I did 80:20 split. Categories are balanced with exception of one category, but I can accept poor performance on one category.
Goal is cloth recognition in images. When I feed image of person wearing pants and tshirt, I want to see two bboxes of these clothes.
My problems:
I already trained few models from tf model zoo. I did over 100k steps on ssd mobilenet v1 and faster rcnn resnet 101.
Problem with ssd is that it won't converge. Loss is not getting below stable 2 and accuracy is bad. Problem with faster rcnn is that loss is below 1 but it's varying a lot and sometimes it jumps over 1.
What I've done:
I tried different batch sizes for ssd with no luck. FRCNN is locked with batch size 1. I improved dataset multiple times. I went from 50 unbalanced classes to 10 balanced classes. I didn't tweak hyperparameters from models configs besides batch size.
My access to strong gpu is limited for me so I can't just randomly try different hyperparams combinations with hope that it will work. Could you suggest me few things that I can do in order to improve my models? I would be very thankful.
machine-learning deep-learning tensorflow convnet object-detection
$endgroup$
bumped to the homepage by Community♦ 1 hour ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
add a comment |
$begingroup$
Situation:
My dataset is 70k images of people wearing clothes. Images are labeled: bbox position and class. There are 10 classes. I did 80:20 split. Categories are balanced with exception of one category, but I can accept poor performance on one category.
Goal is cloth recognition in images. When I feed image of person wearing pants and tshirt, I want to see two bboxes of these clothes.
My problems:
I already trained few models from tf model zoo. I did over 100k steps on ssd mobilenet v1 and faster rcnn resnet 101.
Problem with ssd is that it won't converge. Loss is not getting below stable 2 and accuracy is bad. Problem with faster rcnn is that loss is below 1 but it's varying a lot and sometimes it jumps over 1.
What I've done:
I tried different batch sizes for ssd with no luck. FRCNN is locked with batch size 1. I improved dataset multiple times. I went from 50 unbalanced classes to 10 balanced classes. I didn't tweak hyperparameters from models configs besides batch size.
My access to strong gpu is limited for me so I can't just randomly try different hyperparams combinations with hope that it will work. Could you suggest me few things that I can do in order to improve my models? I would be very thankful.
machine-learning deep-learning tensorflow convnet object-detection
$endgroup$
bumped to the homepage by Community♦ 1 hour ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
add a comment |
$begingroup$
Situation:
My dataset is 70k images of people wearing clothes. Images are labeled: bbox position and class. There are 10 classes. I did 80:20 split. Categories are balanced with exception of one category, but I can accept poor performance on one category.
Goal is cloth recognition in images. When I feed image of person wearing pants and tshirt, I want to see two bboxes of these clothes.
My problems:
I already trained few models from tf model zoo. I did over 100k steps on ssd mobilenet v1 and faster rcnn resnet 101.
Problem with ssd is that it won't converge. Loss is not getting below stable 2 and accuracy is bad. Problem with faster rcnn is that loss is below 1 but it's varying a lot and sometimes it jumps over 1.
What I've done:
I tried different batch sizes for ssd with no luck. FRCNN is locked with batch size 1. I improved dataset multiple times. I went from 50 unbalanced classes to 10 balanced classes. I didn't tweak hyperparameters from models configs besides batch size.
My access to strong gpu is limited for me so I can't just randomly try different hyperparams combinations with hope that it will work. Could you suggest me few things that I can do in order to improve my models? I would be very thankful.
machine-learning deep-learning tensorflow convnet object-detection
$endgroup$
Situation:
My dataset is 70k images of people wearing clothes. Images are labeled: bbox position and class. There are 10 classes. I did 80:20 split. Categories are balanced with exception of one category, but I can accept poor performance on one category.
Goal is cloth recognition in images. When I feed image of person wearing pants and tshirt, I want to see two bboxes of these clothes.
My problems:
I already trained few models from tf model zoo. I did over 100k steps on ssd mobilenet v1 and faster rcnn resnet 101.
Problem with ssd is that it won't converge. Loss is not getting below stable 2 and accuracy is bad. Problem with faster rcnn is that loss is below 1 but it's varying a lot and sometimes it jumps over 1.
What I've done:
I tried different batch sizes for ssd with no luck. FRCNN is locked with batch size 1. I improved dataset multiple times. I went from 50 unbalanced classes to 10 balanced classes. I didn't tweak hyperparameters from models configs besides batch size.
My access to strong gpu is limited for me so I can't just randomly try different hyperparams combinations with hope that it will work. Could you suggest me few things that I can do in order to improve my models? I would be very thankful.
machine-learning deep-learning tensorflow convnet object-detection
machine-learning deep-learning tensorflow convnet object-detection
asked Nov 23 '18 at 19:26
szanksszanks
1
1
bumped to the homepage by Community♦ 1 hour ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
bumped to the homepage by Community♦ 1 hour ago
This question has answers that may be good or bad; the system has marked it active so that they can be reviewed.
add a comment |
add a comment |
1 Answer
1
active
oldest
votes
$begingroup$
Are you training the models from scratch? If yes, then can you try using pre-trained models and fine-tune for your specific dataset.
You'll have the experiment with hyperparameters. Learning rate and optimizer (e.g., sgd, adam, rmsprop, adadelta) will have largest effect on model training and performance.
$endgroup$
add a comment |
Your Answer
StackExchange.ready(function()
var channelOptions =
tags: "".split(" "),
id: "557"
;
initTagRenderer("".split(" "), "".split(" "), channelOptions);
StackExchange.using("externalEditor", function()
// Have to fire editor after snippets, if snippets enabled
if (StackExchange.settings.snippets.snippetsEnabled)
StackExchange.using("snippets", function()
createEditor();
);
else
createEditor();
);
function createEditor()
StackExchange.prepareEditor(
heartbeatType: 'answer',
autoActivateHeartbeat: false,
convertImagesToLinks: false,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: null,
bindNavPrevention: true,
postfix: "",
imageUploader:
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
,
onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
);
);
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f41628%2fhow-to-increase-accuracy-of-model-from-tensorflow-model-zoo%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
1 Answer
1
active
oldest
votes
1 Answer
1
active
oldest
votes
active
oldest
votes
active
oldest
votes
$begingroup$
Are you training the models from scratch? If yes, then can you try using pre-trained models and fine-tune for your specific dataset.
You'll have the experiment with hyperparameters. Learning rate and optimizer (e.g., sgd, adam, rmsprop, adadelta) will have largest effect on model training and performance.
$endgroup$
add a comment |
$begingroup$
Are you training the models from scratch? If yes, then can you try using pre-trained models and fine-tune for your specific dataset.
You'll have the experiment with hyperparameters. Learning rate and optimizer (e.g., sgd, adam, rmsprop, adadelta) will have largest effect on model training and performance.
$endgroup$
add a comment |
$begingroup$
Are you training the models from scratch? If yes, then can you try using pre-trained models and fine-tune for your specific dataset.
You'll have the experiment with hyperparameters. Learning rate and optimizer (e.g., sgd, adam, rmsprop, adadelta) will have largest effect on model training and performance.
$endgroup$
Are you training the models from scratch? If yes, then can you try using pre-trained models and fine-tune for your specific dataset.
You'll have the experiment with hyperparameters. Learning rate and optimizer (e.g., sgd, adam, rmsprop, adadelta) will have largest effect on model training and performance.
answered Nov 24 '18 at 19:46
Brian SpieringBrian Spiering
4,3481129
4,3481129
add a comment |
add a comment |
Thanks for contributing an answer to Data Science Stack Exchange!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
Use MathJax to format equations. MathJax reference.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fdatascience.stackexchange.com%2fquestions%2f41628%2fhow-to-increase-accuracy-of-model-from-tensorflow-model-zoo%23new-answer', 'question_page');
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function ()
StackExchange.helpers.onClickDraftSave('#login-link');
);
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown