Questions tagged [object-detection]

Object detection deals with recognizing the presence of objects of a certain semantic class (e.g., “humans”, “buildings”, “cars”, etc.) in digital image and video data.

Filter by
Sorted by
Tagged with
1
vote
0answers
19 views

YoloV3: Training set, true label boxes vs anchor boxes

I understand the anchor boxes as a set of boxes that based on a comprehensive exploration of the data ( all the true bonding boxes) best describe the variable sizes of all true boxes in the training ...
0
votes
0answers
14 views

How to model the probability of detecting an image, given it is seen multiple times

Are there any existing methods/models describing the probability of an object being detected by a computer vision algorithm given it is seen $n$ times at similar angles and orientations? I know that ...
0
votes
0answers
16 views

What will be the Precision and Recall value for Faster RCNN?

I am using TensorFlow object detection API for Faster RCNN object detector. Now I want to measure the performance of my model, so I have evaluated it using the code below for getting the mAP, ...
0
votes
1answer
21 views

Is there any statistically meaningful definition for object confidence in object detection?

Most modern object detection algorithms rely on neural networks and output a bounding box and confidence for each object (or more accurately, a confidence for each possible object class considered, ...
0
votes
0answers
11 views

How to treat (label and process) border cases in machine learning?

In every computer vision project I struggle with labeling guidelines for border cases. Benchmark datasets don't have this problem, because they are 'cleaned', but in real life unsure cases often ...
2
votes
1answer
93 views

What are “Grids” and Detection at different scales" in YOLOV3?

I've recently started working with Yolov3 and the more I go in depth, the more confused I get. In the simplest terms what I think about YOLOV3 ...
0
votes
0answers
6 views

Can I conclude from object detection metrics on temporal behavior of a detection system?

Assuming I know typical object detection metrics like precision, recall etc. for a certain class, is there anyway to conclude on the probability that an object of that class which is present in ...
0
votes
0answers
7 views

Is cosine distance applicable for multiple classes?

Recently I learnt about DeepSort algorithm which use deep cosine metrics to re-identify objects. However, to my understadning, ...
0
votes
0answers
6 views

Does TensorFlow's Object Detection API models look at the whole image or only the bounded target?

I was wondering if CNNs, specifically the models/feature extractors offered in Tensorflow's Object Detection API, only train on the bounded box of the target image or if it considers the entire image ...
0
votes
0answers
4 views

Evaluating the performance of tracking multiple objects detected with object detection

I have a ground truth dataset where the objects have been manually annotated and each object have been provided an ID that is consistent through time. There are no false positives or false negatives ...
0
votes
1answer
32 views

What is the average precision in the case of no positives for a given category in the context of object detection

In attempting to calculate the average precision of an object detection model, I am wondering about an edge case. Suppose at evaluation time that for a given category, that no detections of that ...
2
votes
1answer
54 views

When computing mAP for an object detection model, how many detections should one consider?

I am trying to write some code to evaluate the MS COCO style mAP (mean average precision, average computed at the category level) at different IOU levels in the context of object detection with a ...
0
votes
0answers
38 views

Interpretation of Precision and Recall from Object detection API?

In given picture I have precision and recall value -1.000. What does this signify? Could someone please help me to interpret this results? Furthermore Can I calculate F1 score for this average ...
0
votes
0answers
10 views

What are the metrics to deciede if a dataset is good enough for the deep learning task of object detection task?

I am asked by my supervisor to identify certain metrics with which we could definitively say that a dataset is suitable for the deep learning task of object detection.I did some research and found ...
0
votes
0answers
4 views

pretrained model question

I'm quite confuse about the pretrained model. Let say I'm training the rcnn or cascade rcnn. They have the backbone for classification. like resnet 50 or resnet 101 If I want to make a pretrained ...
0
votes
0answers
60 views

Two True Positive for one ground truth in object detection

I am wondering is it possible to have two true positive predictions for one bounding box ground truth only. Following this section from Stanford. They define truth positive like this: We start with ...
1
vote
0answers
42 views

Enemy detection in 3D video games without labeled dataset

I'm working on a tool for single-class object detection in video games, which should be able to detect enemies in 3D environments. The main problem is, that I cannot use any labeled dataset. Instead I ...
0
votes
0answers
14 views

Object detection vs segmentation?

My problem statement is as follows: "Object detection is the concept of classifiying & localising an object in an image, and semantic segmentation is the concept of labeling each pixel to a ...
0
votes
0answers
15 views

Evaluate result of a Detection algorithm based on the object's shape as Ground truth

I'm working on project where i have to detect small colored cars driving on rails from static camera (Bird's Eye-view see below image) The algorithm in short outputs first a mask image with only the ...
0
votes
0answers
8 views

How many Test-videos are needed to evaluate a detection algorithm?

I've developed an object-detection algorithm based on combinations of available approaches in the literature (using classical approches). The algorithm is designed for a special application which ...
0
votes
1answer
32 views

Resize image in object detection task of computer vision

In object detection, they usually resize by keeping the ratio the same as the original image, which usually names "letterbox" resize. My question is: Why we need to do that? As I see with ...
0
votes
0answers
15 views

What activation and where to use in MaskRCNN RPN

so I've been trying to implement my own version of MaskRCNN, and I am baffled by how the RPN is implemented in various places. Assuming the standard RPN architecture of a shared 3x3 Conv2d, and two ...
0
votes
0answers
81 views

Faster RCNN for one class in object detection

Let say I have a task to detect the bounding box of one object only. And the only thing I care about is the IoU between prediction and ground truth, no need for real-time. My question: Should I ...
0
votes
1answer
162 views

How many images with bounding boxes are in the ImageNet object detection dataset?

I am trying to understand how many images with object-detection bounding boxes are in ImageNet dataset, and how many objects are there in these images. In MS-COCO paper, it says: "Recently, a ...
0
votes
0answers
33 views

Should I keep training samples without labels

I want to supervisely train an object detector for pedestrians, this detector is expected to ideally detect all the pedestrians in an image, and output the detected pedestrians as 2D bounding boxes, ...
0
votes
1answer
246 views

how to find mean average precision of object detection algorithms

To start with, I would like to mention another question which was asked in a better way. But my problem differs. pseudo code for the algorithms I have four different object detection algorithms which ...
1
vote
0answers
28 views

Why Anchor Boxes in Object Detection

What I understand from the anchor boxes is that they are predefined set of boxes of different sizes and shapes. A model like YOLO predicts offsets of the coordinates for each anchor box along with ...
1
vote
0answers
81 views

How to measure the performance of Mask RCNN model. Given that there are two tasks , one object detection and another image segmentation

Mask RCNN is an instance image segmentation technique. It is based on Faster RCNN for object detection and an additional mask operation is performed by another CNN.
2
votes
1answer
645 views

1 neuron BCE loss VS 2 neurons CE loss

I built a custom version of YOLO that should only detect one type of objects, where the objectness measure (which tells how likely a bounding-box contains an object of any type), is learned using a ...
0
votes
0answers
11 views

TF object detection - The total number of detected objects is not increasing

I'm building a model to recognize fishes in the aquarium (150 different fishes). I'm using a faster_rcnn_inception_v2_coco_2018_01_28 model for transfer learning from TF object detection API. I have ...
1
vote
0answers
41 views

Object Classification vs Object Detection

My problem statement is as follows: I've got some technical documents as JPEGs, mostly text, but with some standard pictograms. Each image does not need to have all the pictograms or any pictogram at ...
1
vote
0answers
52 views

yolo v1: how gradients in grids with no objects overpower

I am reading yolo v1 paper. I am having confusion while trying to understand this: "Also, in every image many grid cells do not contain any object. This pushes the “confidence” scores of those ...
0
votes
0answers
19 views

Does the selective search algorithm in object detection learn?

I am trying to get a better grasp of how object detection works. I (almost) completely understand the concept behind RPNs. However I am little bit confused with the selective search algorithm part. ...
0
votes
1answer
121 views

yolo v1: why we weigh localization error more than classification error

I'm reading yolo v1 paper. I am trying to understand this part of the paper: "We use sum-squared error because it is easy to optimize, however it does not perfectly align with our goal of ...
0
votes
0answers
28 views

Yolo for object detection

I am working on (RSNA Pneumonia Detection Challenge) of kaggle. In which we have to find if a patient is suffering from pneumonia and if suffering then find the abnormality in the image by finding ...
1
vote
1answer
43 views

Which Object detection model will give the best result on images when the speed is not a problem for Text Images

I want to develop a model for cropping the equations from the Maths questions as people like me are struggling a lot for doing it manually for the research purpose. I want to know if we can do this? ...
0
votes
0answers
20 views

Faster R-CNN : How to change max pool to average pool in pre-trained tensor-flow object detection model

I am using pre-trained faster R-CNN tensor flow object detection API for my use case. I want to change ROI pooling layer from existing max pool to average pool, how can I do that. Is it possible to do ...
0
votes
0answers
9 views

How to evaluate (compare) probmaps?

In the field of Object Detection where bounding boxes is the output I suppose comparing results is easy, but with the problem of Road detection, in which the output is usually Probmaps, how do you ...
0
votes
0answers
106 views

Why is SSD object detection model overfitting even with very high weight decay?

Actually I am struggling for a long time with this problem and had tried a lot of experiments. I am working on an object detection model using the SSD architecture with various backbones. I started ...
1
vote
0answers
37 views

Estimating epistemic uncertainty in object detection

I've begun to explore a dataset wherein I am determining whether an object is present, after making N attempted indep. observations of the same object's position. While N attempts may be made, the ...
0
votes
0answers
63 views

Can I use the sigmoid activation function to predict bounding box coordinates?

I’m designing a network which localizes images of individual fruit belonging to one of six different classes of fruit. The target bounding box parameters are normalized so that they assume values ...
0
votes
0answers
9 views

What is the difference between HLC (Histogram of local features) , CSS ( color self-similarity) ans MDST (Max DisSimilarity of Different Templates)

I'm new to computer vision and have been researching for Master thesis purposes in Detection algorithms and the techniques used in each. As I arrived to the point where alot of papers showed the ...
0
votes
0answers
14 views

How deep does the following convolutional neural network need to be?

Over the course of the last two days, I’ve trained two convolutional neural networks, The first of these networks comprised of 3 sets of one convolutional layer, followed by a max pooling layer, and ...
0
votes
1answer
19 views

Do we need to back propagate through non-max suppression layers?

In convolutional networks designed to serve the purpose of object detection, following the output of the final activations in the network, said activations are processed in accordance with the theory ...
0
votes
0answers
45 views

Need advice improving object detection model

i have 433 image dataset which consists of class (multiple class can be inside one image): 'a': 634, 'b': 501, 'c': 590, 'd': 293, 'e': 524, 'f': 262 i know my ...
0
votes
0answers
50 views

Equivalence of maP and confusion matrix in object detection

I have a piece of code that computes maP for object detection using this approach I have another piece of code that computes a confusion matrix for object detection using this approach If I then use ...
2
votes
0answers
424 views

Resize images before training object detection

I am training an object detector. I didn't resize my image before labeling because the of assumption that the model does this automatically to fit its input shape. ...
0
votes
0answers
6 views

Will running many classification models sequentially be similar to running only 1 detection model?

This question is related to this other one: In what situation many classification models (specialized) will be better to use instead of just 1 (generalist)? With 1 classification model it is possible ...
0
votes
0answers
15 views

In what situation many classification models (specialized) will be better to use instead of just 1 (generalist)?

Recently, I started working on Classification and Object detection. I am using a dev board to make ONLY inference. So I am just using the models created by someone else. As I have to develop a demo, ...
0
votes
1answer
17 views

Invoice Extraction

What topics should I read on to prepare myself for data extraction from invoices (in image file or PDF). Basically I will have a file with invoice and need to extract from it such information as ...