Auto Topic: detector
auto_detector | topic
Coverage Score
1
Mentioned Chunks
15
Mentioned Docs
1
Required Dimensions
definitionpros_cons
Covered Dimensions
definitionpros_cons
Keywords
detector
Relations
| Source | Type | Target | W |
|---|---|---|---|
| Auto Topic: detector | CO_OCCURS | Auto Topic: window | 7 |
| Auto Topic: detector | CO_OCCURS | Auto Topic: pixels | 6 |
| Auto Topic: detector | CO_OCCURS | Auto Topic: windows | 4 |
| Auto Topic: convolutional | CO_OCCURS | Auto Topic: detector | 3 |
| Auto Topic: cnn | CO_OCCURS | Auto Topic: detector | 3 |
Evidence Chunks
| Source | Confidence | Mentions | Snippet |
|---|---|---|---|
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.61 | 4 | ... Each location receives inputs from pixels in a window about that location. The output of the ReLU, as we have seen, forms a simple pattern detector. Now if we put a second layer on top of this, each location in the second layer receives inputs from first-layer values in a window ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.61 | 4 | ... closer to the horizon in the image are farther away from the camera, and so must be smaller in the image. This means we can rule out some detector responses—if a detector finds a pedestrian who is large in the image and whose feet are close to the horizon, it has found an enormou ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.59 | 3 | to bottom): a horizontal bar detector; a vertical bar detector; and (harder to note) a line ending detector. These detectors pay attention to the contrast of the bar, so (for example) a horizontal bar that is light on top and dark below produces a positive (green) response, and o ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.59 | 3 | rong. In turn, a reasonably reliable pedestrian detector is capable of producing estimates of the horizon, if there are several pedestrians in the scene at different distances from the camera. This is because the relative scaling of the pedestrians is a cue to where the horizon i ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.59 | 3 | ... d. For the original images see Aneja et al. (2018). people do so many things in so many contexts. For example, suppose we have a pedestrian detector that performs well on a large data set. There will be rare phenomena (for example, people mounting unicycles) that do not appear in ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... larm as parent. • Adding Earthquake: If the alarm is on, it is more likely that there has been an earth- quake. (The alarm is an earthquake detector of sorts.) But if we know that there has been a burglary, then that explains the alarm, and the probability of an earthquake would ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... actice. (For example, it is not easy to imagine how a model-free approach would enable one to design and build, say, the LIGO gravity- wave detector.) Our intuition, for what it’s |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... actice. (For example, it is not easy to imagine how a model-free approach would enable one to design and build, say, the LIGO gravity- wave detector.) Our intuition, for what it’s worth, is that as the environment becomes more complex, the advantages of a model-based approach bec ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... onvolutional neural networks do well. You should think of a layer—a con- volution followed by a ReLU activation function—as a local pattern detector (Figure 27.12). The convolution measures how much each local window of the image looks like the kernel pattern; the ReLU sets low-s ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... ect.1 The set ofBounding box classes is fixed in advance. So we might try to detect all faces, all cars, or all cats. We can build an object detector by looking at a small sliding window onto the largerSliding window image—a rectangle. At each spot, we classify what we see in the ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... haps only the objects that appear large in the image—the front row—should be reported. The choice depends on the intended use of the object detector. • Report precise locations of objects using these windows: Once we know that the object is somewhere in the window, we can afford ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... ess test. A network that finds regions with objects is called a regional proposal network (RPN). Regional proposal network (RPN) The object detector known as Faster RCNN encodes a large collection of bounding boxes as a map of fixed size. Then it builds a network that can predict a ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... a ground truth category label and bounding box. Usually, the boxes and labels are supplied by humans. Then we feed each image to the object detector and compare its output to the ground truth. We should be willing to accept boxes that are off by a few pixels, because the ground t ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... eural net classifier 0.9 0.7 Figure 27.13 Faster RCNN uses two networks. A picture of a young Nelson Mandela is fed into the object detector. One network computes “objectness” scores of candidate image boxes, called “anchor boxes,” centered at a grid point. There are nine anchor ... |
textbook Artificial-Intelligence-A-Modern-Approach-4th-Edition.pdf | 0.55 | 1 | ... because they rely so strongly on context. For example, a classifier that labels “swimming” sequences very well might just be a swimming pool detector, which wouldn’t work for (say) swimmers in rivers. More general problems remain open—for example, how to link observations of the b ... |