An Efficient Multiscale Pyramid Attention Network for Face Detection in Surveillance Images

<table class="table-group" id="tab2"><tr><td><table class="table"><tr><td class="thead-hr" colspan="4"><hr/></td></tr><tr class="thead"><td class="align_left">Layer</td><td class="align_center">Stride</td><td class="align_center">Anchor size (pixels)</td><td class="align_center">Boxes</td></tr><tr><td class="thead-hr" colspan="4"><hr/></td></tr><tr><td class="align_left">Conv3_3</td><td class="align_center">4</td><td class="align_center">16/ <span class="nowrap"><svg height="13.1484pt" id="M18" style="vertical-align:-1.3645pt" version="1.1" viewbox="-0.0498162 -11.7839 16.3323 13.1484" width="16.3323pt" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"><g transform="matrix(.013,0,0,-0.013,0,-1.543)"><path d="M782 731H740L476 -79H474L280 352L133 274L151 239L242 286L473 -219L782 731Z"></path></g><rect height="0.65243" width="6.26339" x="9.96925" y="-11.0816"></rect><g transform="matrix(.013,0,0,-0.013,9.969,0)"><path d="M412 140C382 77 369 73 315 73H129L270 222C362 320 402 379 402 466C402 571 322 635 234 635C177 635 130 609 99 576L42 495L64 475C90 514 133 568 201 568C274 568 318 519 318 435C318 349 255 267 193 193C144 135 87 78 32 23V0H405C417 45 427 89 440 131L412 140Z"></path></g></svg>,</span> 16</td><td class="align_center">32768</td></tr><tr><td class="align_left">Conv4_3</td><td class="align_center">8</td><td class="align_center">32/ <span class="nowrap"><svg height="13.1484pt" id="M19" style="vertical-align:-1.3645pt" version="1.1" viewbox="-0.0498162 -11.7839 16.3323 13.1484" width="16.3323pt" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"><g transform="matrix(.013,0,0,-0.013,0,-1.543)"><path d="M782 731H740L476 -79H474L280 352L133 274L151 239L242 286L473 -219L782 731Z"></path></g><rect height="0.65243" width="6.26339" x="9.96925" y="-11.0816"></rect><g transform="matrix(.013,0,0,-0.013,9.969,0)"><path d="M412 140C382 77 369 73 315 73H129L270 222C362 320 402 379 402 466C402 571 322 635 234 635C177 635 130 609 99 576L42 495L64 475C90 514 133 568 201 568C274 568 318 519 318 435C318 349 255 267 193 193C144 135 87 78 32 23V0H405C417 45 427 89 440 131L412 140Z"></path></g></svg>,</span> 32</td><td class="align_center">8192</td></tr><tr><td class="align_left">Conv5_3</td><td class="align_center">16</td><td class="align_center">64/ <span class="nowrap"><svg height="13.1484pt" id="M20" style="vertical-align:-1.3645pt" version="1.1" viewbox="-0.0498162 -11.7839 16.3323 13.1484" width="16.3323pt" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"><g transform="matrix(.013,0,0,-0.013,0,-1.543)"><path d="M782 731H740L476 -79H474L280 352L133 274L151 239L242 286L473 -219L782 731Z"></path></g><rect height="0.65243" width="6.26339" x="9.96925" y="-11.0816"></rect><g transform="matrix(.013,0,0,-0.013,9.969,0)"><path d="M412 140C382 77 369 73 315 73H129L270 222C362 320 402 379 402 466C402 571 322 635 234 635C177 635 130 609 99 576L42 495L64 475C90 514 133 568 201 568C274 568 318 519 318 435C318 349 255 267 193 193C144 135 87 78 32 23V0H405C417 45 427 89 440 131L412 140Z"></path></g></svg>,</span> 64</td><td class="align_center">2048</td></tr><tr><td class="align_left">Conv6_2</td><td class="align_center">32</td><td class="align_center">128/ <span class="nowrap"><svg height="13.1484pt" id="M21" style="vertical-align:-1.3645pt" version="1.1" viewbox="-0.0498162 -11.7839 16.3323 13.1484" width="16.3323pt" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"><g transform="matrix(.013,0,0,-0.013,0,-1.543)"><path d="M782 731H740L476 -79H474L280 352L133 274L151 239L242 286L473 -219L782 731Z"></path></g><rect height="0.65243" width="6.26339" x="9.96925" y="-11.0816"></rect><g transform="matrix(.013,0,0,-0.013,9.969,0)"><path d="M412 140C382 77 369 73 315 73H129L270 222C362 320 402 379 402 466C402 571 322 635 234 635C177 635 130 609 99 576L42 495L64 475C90 514 133 568 201 568C274 568 318 519 318 435C318 349 255 267 193 193C144 135 87 78 32 23V0H405C417 45 427 89 440 131L412 140Z"></path></g></svg>,</span> 128</td><td class="align_center">512</td></tr><tr><td class="align_left">Conv7_2</td><td class="align_center">64</td><td class="align_center">256/ <span class="nowrap"><svg height="13.1484pt" id="M22" style="vertical-align:-1.3645pt" version="1.1" viewbox="-0.0498162 -11.7839 16.3323 13.1484" width="16.3323pt" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink"><g transform="matrix(.013,0,0,-0.013,0,-1.543)"><path d="M782 731H740L476 -79H474L280 352L133 274L151 239L242 286L473 -219L782 731Z"></path></g><rect height="0.65243" width="6.26339" x="9.96925" y="-11.0816"></rect><g transform="matrix(.013,0,0,-0.013,9.969,0)"><path d="M412 140C382 77 369 73 315 73H129L270 222C362 320 402 379 402 466C402 571 322 635 234 635C177 635 130 609 99 576L42 495L64 475C90 514 133 568 201 568C274 568 318 519 318 435C318 349 255 267 193 193C144 135 87 78 32 23V0H405C417 45 427 89 440 131L412 140Z"></path></g></svg>,</span> 256</td><td class="align_center">128</td></tr><tr class="table-tr"><td colspan="4"><hr class="tbody-hr"/></td></tr></table></td></tr></table>

<div>Parameters of the five anchors for various sizes of surveillance face detection.</div>

Security and Communication Networks

tab2

Table 2

Table 2: An Efficient Multiscale Pyramid Attention Network for Face Detection in Surveillance Images 