py-faster-rcnn | Faster R-CNN ( Python implementation
kandi X-RAY | py-faster-rcnn Summary
kandi X-RAY | py-faster-rcnn Summary
Faster R-CNN (Python implementation) -- see for the official MATLAB version
Support
Quality
Security
License
Reuse
Top functions reviewed by kandi - BETA
- Return the roidb dictionary
- Return roidb handler
- Call build extensions
- Set the cuda compiler to use
- Load configuration from file
- Recursively merge a b into b
- Locate CUDA
- Find a file in a search path
- Append flipped images
- Returns the widths of the image
- Setup the image
- Reshape the object
- Forward the pixel to the given bottom
- Create a config dictionary from a list
- Get an imdb dataset
- Adds a path to sys path
- Turn on
- Set proposal method
- Parse arguments
py-faster-rcnn Key Features
py-faster-rcnn Examples and Code Snippets
"Unknown layer type: Python"
WITH_PYTHON_LAYER := 1
cd $ROOT_DIR/py-faster-rcnn/caffe-fast-rcnn/
make clean
make -j8 && make pycaffe
fatal error: caffe/proto/caffe.pb.h: No such file or directory
#include "caffe/proto/caffe.pb.h"
layer {
name: "loss_cls_cas_128"
type: "SoftmaxWithCascadeLoss"
bottom: "cls_score_cas_128"
bottom: "labels"
bottom: "bp_map_cas_128"
top: "loss_cls_cas_128"
top: "bp_map_cas_64"
loss_weight: 1
loss_weight: 0
loss_param {
hard
optional ROIPoolingParameter roi_pooling_param = 200;
optional SmoothL1LossParameter smooth_l1_loss_param = 201;
// Message that stores parameters used by ROIPoolingLayer
message ROIPoolingParameter {
// Pad, kernel size, and stride are all g
def ap_per_class(tp, conf, pred_cls, target_cls):
""" Compute the average precision, given the recall and precision curves.
Method originally from https://github.com/rafaelpadilla/Object-Detection-Metrics.
# Arguments
tp: True
def compute_ap(recall, precision):
""" Compute the average precision, given the recall and precision curves.
Code originally from https://github.com/rbgirshick/py-faster-rcnn.
# Arguments
recall: The recall curve (list).
Community Discussions
Trending Discussions on py-faster-rcnn
QUESTION
Getting the nvcc fatal : '--ptxas-options=-v': expected a number
error when I try to build a Windows port of Faster-RCNN. You may reach the setup file (which is a Python script) directly from here.
Software Environment:
...ANSWER
Answered 2019-Jun-09 at 14:07This configuration line is no longer correct with CUDA 10.1:
QUESTION
ANSWER
Answered 2018-Dec-28 at 16:47I don’t know what is happening in the first case, but in the second case looks like the left and right locations are both inside the bounding box. The number of pixels spanned must therefore include both left and right locations. This is why 1 is added.
QUESTION
I got this sentence in py-faster-Rcnn
...ANSWER
Answered 2018-Feb-07 at 07:56Install CUDA 8.0 and that should resolve the issue.
From this you can find out that GeForce GTX 1080 Ti GPU belongs to Pascal GPU Micro architecture with compute capability version 6.1 which is supported by CUDA 8.0 or higher.
CUDA SDK 8.0 support for compute capability 2.0 – 6.x (Fermi, Kepler, Maxwell, Pascal), last version with support for Compute Capability 2.x (Fermi)
CUDA SDK 7.5 support for compute capability 2.0 – 5.x (Fermi, Kepler, Maxwell)
QUESTION
I am using Google API for object detection in tensorflow to train and infer on a custom dataset.
I would like to adjust the parameters of the config file to better suit my samples (e.g. no. of region proposals, size of ROI bbox, etc.). To do so, I need to know what each parameter does. Unfortunately, the config files (found here ) do not have comments or explanations. Some, such as "num classes" are self-explanatory, but others are tricky.
I found this file with more comments , but wasn't able to 'translate' it to my format.
I would like to know one of the following: 1. explanation of each parameter for google's API config file or 2. 'translation' from the official faster-rcnn to google's API config or at least 3. thorough review of faster-rcnn with technical details of the parameters (the official article doesn't provide all the details)
Thank you for your kind help !
Example of a config file:
...ANSWER
Answered 2018-Jan-23 at 15:56I found two sources that shed some light on the config file: 1. The folder protos inside tensorflow github covers all configuration options with some comments on each options. You should checkout faster_rcnn.proto , eval.proto and train.proto for the most common 2. This blog post by Algorithmia covers thoroughly all steps to download, prepare and train faster RCNN on Google's Open Images dataset. 2/3-way through, there is some discussion on the configuration options.
QUESTION
I am looking through the code in the library. In the paper, (page 6, second column, first para), it is stated as convolutional layers are fixed (during training in third and fourth steps) and tuned the RPN layers and Fast RCNN layers.
Which portion of the code is taking care of it?
I looked at the code and Solver.cpp is the one controlling Forward/Backward. I don't see implementation of fixing convolutional layers there.
Then all prototxt files have similar implementations for layers.
How this fixing convolutional layers in training is implemented?
...ANSWER
Answered 2017-Dec-13 at 06:29When freezing a layer during fine-tuning, one usually sets
QUESTION
I am trying to debug C++ code behind python for this library (faster rcnn).
I followed the instruction here.
Say I like to stop a break point at solver.cpp and line 188.
...ANSWER
Answered 2017-Dec-12 at 06:25You probably has passed your breakpoint (before GDB got started). You should, per the instructions mentioned:
run python under gdb from the start
Try to run
your python
under GDB.
Read about Debugging with GDB.
You might want to use a python
interpreter built with DWARF debug info, so compiled with -g
(and of course your solver.cpp
should also be compiled with -g
). Since Python is free software recompiling it with -g
should be pretty simple.
BTW, are you sure of the line position 188 of your breakpoint? Did you try to add more breakpoints in solver.cpp
at other positions, or to break at function names?
QUESTION
I use Faster RCNN library for Deep Learning and here is a discussion of how to train own dataset. That is one step ahead of me.
For me I like to understand how dataset for training is setup and how it is loaded for training?
When I look at the code, I saw this line of code
imdb = get_imdb(imdb_name)
from train_faster_rcnn_alt_opt.py
and it calls factory.py inside datasets folder
.
factory.py has
...ANSWER
Answered 2017-Nov-23 at 07:22Now I understood. lib/datasets/factory.py has
QUESTION
I use the py-faster-rcnn demo to build further of my project with 20 classes. However, I am trying to gain the softmax, last layer probability of my classes.
For example:
...ANSWER
Answered 2017-Nov-21 at 14:35The raw scores
the detector outputs include overlapping detections and very low score detections as well.
Note that only after applying non-maximal suppression (aka "nms") with NMS_THRESH=0.3
the function vis_detection
only displays detections with confidence larger than CONF_THRESH=0.8
.
So, if you want to look at the "true" objects, you need to check inside vis_detection
and check only the detections it renders on the image.
QUESTION
I am testing Faster Rcnn. Installation is fine. During the installation, I had one issue with cudnn5.1 and I followed the suggestion here and now installation is fine.
Now I test the demo code as
...ANSWER
Answered 2017-Nov-17 at 04:38Now it works for me. Since libcudnn5_5.1 is for CUDA7.5. Can check in cudnn's user guide at GPU and driver requirements. So I changed to cudnnv6.0 for CUDA8.0.
Then you may face the issue of
QUESTION
I understand the normal lambda expression, such as
...ANSWER
Answered 2017-Sep-23 at 04:20A lambda function:
Community Discussions, Code Snippets contain sources that include Stack Exchange Network
Vulnerabilities
No vulnerabilities reported
Install py-faster-rcnn
Clone the Faster R-CNN repository
We'll call the directory that you cloned Faster R-CNN into FRCN_ROOT Ignore notes 1 and 2 if you followed step 1 above. Note 1: If you didn't clone Faster R-CNN with the --recursive flag, then you'll need to manually clone the caffe-fast-rcnn submodule: git submodule update --init --recursive Note 2: The caffe-fast-rcnn submodule needs to be on the faster-rcnn branch (or equivalent detached state). This will happen automatically if you followed step 1 instructions.
Build the Cython modules cd $FRCN_ROOT/lib make
Build Caffe and pycaffe cd $FRCN_ROOT/caffe-fast-rcnn # Now follow the Caffe installation instructions here: # http://caffe.berkeleyvision.org/installation.html # If you're experienced with Caffe and have all of the requirements installed # and your Makefile.config in place, then simply do: make -j8 && make pycaffe
Download pre-computed Faster R-CNN detectors cd $FRCN_ROOT ./data/scripts/fetch_faster_rcnn_models.sh This will populate the $FRCN_ROOT/data folder with faster_rcnn_models. See data/README.md for details. These models were trained on VOC 2007 trainval.
Pre-trained ImageNet models can be downloaded for the three networks described in the paper: ZF and VGG16. VGG16 comes from the Caffe Model Zoo, but is provided here for your convenience. ZF was trained at MSRA.
Support
Reuse Trending Solutions
Find, review, and download reusable Libraries, Code Snippets, Cloud APIs from over 650 million Knowledge Items
Find more librariesStay Updated
Subscribe to our newsletter for trending solutions and developer bootcamps
Share this Page