Comments (5)
Can you try to run the provided example/ssd/score_ssd_pascal.py and see if there are results saved in the output directory?
Make sure you have provided the valid output_directory, label_map_file, and name_size_file according to here.
Please let me know what you find.
BTW. The mAP looks very high. If you didn't mix testing images with training, then congratulations :)
from caffe.
Actually, i have run the provided example/ssd/score_ssd_pascal.py successfully. And the results (txt files, which are blank) are saved. I'm sure i have provided valid output_directory, label_map_file, and name_size_file. To make my confusion clearly: I have obtained the output files, but all those files are blank...
Thanks for your reminder about the mAP issue. my data is small, only have 5 classes object. And the number of test images is only 119, (training data num. is 1084). So this may be the reason why mAP is very high.
from caffe.
Ah, you should change this line to:
test_iter = math.ceil(float(num_test_image) / test_batch_size)
or you could change test_batch_size to 1.
The reason is that 119 / 8 = 14, and it will never reach to here.
from caffe.
Wow, thank god i have mentioned the number 119...Also, thank you very much for detailed help! I have done with this problem!
BTW,
test_iter = math.ceil(float(num_test_image) / test_batch_size)
should be added an int type transform or it will lead to an error about data type.
test_iter = int(math.ceil(float(num_test_image) / test_batch_size))
Again, thank you for your help~
from caffe.
You are right. I updated the script and added a note to it. Thanks for pointing out the bug!
from caffe.
Related Issues (20)
- 您这个版本的caffe 对比 BVLC/caffe 都做了哪些改动?
- Dataset file VOCtest_06-Nov-2007.tar download failed,anyone can give me a new download link?
- Problems of port_def.inc file not found in Make compilation HOT 1
- Where I can find image name when I train model?
- Single Shot Detector (SSD) architecture modification option?
- [libprotobuf ERROR google/protobuf/text_format.cc:288] Error parsing text-format caffe.NetParameter: 1761:15: Message type "caffe.PriorBoxParameter" has no field named "fixed_size".
- caffe.layerparameter has no field named norm_param HOT 1
- How to evaluate results after running train.sh??
- Compilation Error in cudnn_conv_layer.cpp: 'CUDNN_CONVOLUTION_FWD_SPECIFY_WORKSPACE_LIMIT’ was not declared in this scope HOT 8
- caffe ssd build failed layers/detection_output_layer.o' failed HOT 1
- Error while building Caffe - SSD tree
- Which data in trainval.txt is set aside for validation??
- Unable to download fully convolutional reduced VGGNet HOT 1
- Cannot compile Caffe (cannot find -lhdf5, cannot find -lhdf5) HOT 1
- build/tools/get_image_size: No such file or directory HOT 1
- Do annotated data layer include to rotate image method?
- Loss and missing detections
- Check failed: error == cudaSuccess (8 vs. 0) on the nvidia-docker2, RTX3070
- Priorboxes sizes and aspect ratios
- The cloud drive link provided by the author for the pre-trained model cannot be found.
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from caffe.