1 Star 0 Fork 18

sunsonzh/Mobilenet-SSD-License-Plate-Detection

加入 Gitee
与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :)
免费加入
文件
该仓库未声明开源许可证文件(LICENSE),使用请关注具体项目描述及其代码上游依赖。
克隆/下载
mssd512_voc.prototxt 26.39 KB
一键复制 编辑 原始数据 按行查看 历史
jack_yu_ 提交于 2019-07-22 18:11 . Add files via upload
1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677787980818283848586878889909192939495969798991001011021031041051061071081091101111121131141151161171181191201211221231241251261271281291301311321331341351361371381391401411421431441451461471481491501511521531541551561571581591601611621631641651661671681691701711721731741751761771781791801811821831841851861871881891901911921931941951961971981992002012022032042052062072082092102112122132142152162172182192202212222232242252262272282292302312322332342352362372382392402412422432442452462472482492502512522532542552562572582592602612622632642652662672682692702712722732742752762772782792802812822832842852862872882892902912922932942952962972982993003013023033043053063073083093103113123133143153163173183193203213223233243253263273283293303313323333343353363373383393403413423433443453463473483493503513523533543553563573583593603613623633643653663673683693703713723733743753763773783793803813823833843853863873883893903913923933943953963973983994004014024034044054064074084094104114124134144154164174184194204214224234244254264274284294304314324334344354364374384394404414424434444454464474484494504514524534544554564574584594604614624634644654664674684694704714724734744754764774784794804814824834844854864874884894904914924934944954964974984995005015025035045055065075085095105115125135145155165175185195205215225235245255265275285295305315325335345355365375385395405415425435445455465475485495505515525535545555565575585595605615625635645655665675685695705715725735745755765775785795805815825835845855865875885895905915925935945955965975985996006016026036046056066076086096106116126136146156166176186196206216226236246256266276286296306316326336346356366376386396406416426436446456466476486496506516526536546556566576586596606616626636646656666676686696706716726736746756766776786796806816826836846856866876886896906916926936946956966976986997007017027037047057067077087097107117127137147157167177187197207217227237247257267277287297307317327337347357367377387397407417427437447457467477487497507517527537547557567577587597607617627637647657667677687697707717727737747757767777787797807817827837847857867877887897907917927937947957967977987998008018028038048058068078088098108118128138148158168178188198208218228238248258268278288298308318328338348358368378388398408418428438448458468478488498508518528538548558568578588598608618628638648658668678688698708718728738748758768778788798808818828838848858868878888898908918928938948958968978988999009019029039049059069079089099109119129139149159169179189199209219229239249259269279289299309319329339349359369379389399409419429439449459469479489499509519529539549559569579589599609619629639649659669679689699709719729739749759769779789799809819829839849859869879889899909919929939949959969979989991000100110021003100410051006100710081009101010111012101310141015101610171018101910201021102210231024102510261027102810291030103110321033103410351036103710381039104010411042104310441045104610471048104910501051105210531054105510561057105810591060106110621063106410651066106710681069107010711072107310741075107610771078107910801081108210831084108510861087108810891090109110921093109410951096109710981099110011011102110311041105110611071108110911101111111211131114111511161117111811191120112111221123112411251126112711281129113011311132113311341135113611371138113911401141114211431144114511461147114811491150115111521153115411551156115711581159116011611162116311641165116611671168116911701171117211731174117511761177117811791180118111821183118411851186118711881189119011911192119311941195119611971198119912001201120212031204120512061207120812091210121112121213121412151216121712181219122012211222122312241225122612271228122912301231123212331234123512361237123812391240124112421243124412451246124712481249125012511252125312541255125612571258125912601261126212631264126512661267126812691270127112721273127412751276127712781279128012811282128312841285128612871288128912901291129212931294129512961297129812991300130113021303130413051306130713081309131013111312131313141315131613171318131913201321132213231324132513261327132813291330133113321333133413351336133713381339134013411342134313441345134613471348134913501351135213531354135513561357135813591360136113621363136413651366136713681369137013711372137313741375137613771378137913801381138213831384138513861387138813891390139113921393139413951396139713981399140014011402140314041405140614071408140914101411141214131414141514161417141814191420142114221423142414251426142714281429143014311432143314341435143614371438143914401441144214431444144514461447144814491450145114521453145414551456145714581459146014611462
layer {
name: "data"
type: "Input"
top: "data"
input_param {
shape {
dim: 1
dim: 3
dim: 512
dim: 512
}
}
}
layer {
name: "ssd0_mobilenet0_conv0"
type: "Convolution"
bottom: "data"
top: "ssd0_mobilenet0_conv0"
convolution_param {
num_output: 8
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm0/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv0"
top: "ssd0_mobilenet0_conv0"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv1"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv0"
top: "ssd0_mobilenet0_conv1"
convolution_param {
num_output: 8
bias_term: true
pad: 1
kernel_size: 3
group: 8
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm1/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv1"
top: "ssd0_mobilenet0_conv1"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv2"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv1"
top: "ssd0_mobilenet0_conv2"
convolution_param {
num_output: 16
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm2/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv2"
top: "ssd0_mobilenet0_conv2"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv3"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv2"
top: "ssd0_mobilenet0_conv3"
convolution_param {
num_output: 16
bias_term: true
pad: 1
kernel_size: 3
group: 16
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm3/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv3"
top: "ssd0_mobilenet0_conv3"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv4"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv3"
top: "ssd0_mobilenet0_conv4"
convolution_param {
num_output: 32
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm4/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv4"
top: "ssd0_mobilenet0_conv4"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv5"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv4"
top: "ssd0_mobilenet0_conv5"
convolution_param {
num_output: 32
bias_term: true
pad: 1
kernel_size: 3
group: 32
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm5/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv5"
top: "ssd0_mobilenet0_conv5"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv6"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv5"
top: "ssd0_mobilenet0_conv6"
convolution_param {
num_output: 32
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm6/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv6"
top: "ssd0_mobilenet0_conv6"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv7"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv6"
top: "ssd0_mobilenet0_conv7"
convolution_param {
num_output: 32
bias_term: true
pad: 1
kernel_size: 3
group: 32
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm7/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv7"
top: "ssd0_mobilenet0_conv7"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv8"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv7"
top: "ssd0_mobilenet0_conv8"
convolution_param {
num_output: 64
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm8/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv8"
top: "ssd0_mobilenet0_conv8"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv9"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv8"
top: "ssd0_mobilenet0_conv9"
convolution_param {
num_output: 64
bias_term: true
pad: 1
kernel_size: 3
group: 64
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm9/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv9"
top: "ssd0_mobilenet0_conv9"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv10"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv9"
top: "ssd0_mobilenet0_conv10"
convolution_param {
num_output: 64
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm10/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv10"
top: "ssd0_mobilenet0_conv10"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv11"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv10"
top: "ssd0_mobilenet0_conv11"
convolution_param {
num_output: 64
bias_term: true
pad: 1
kernel_size: 3
group: 64
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm11/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv11"
top: "ssd0_mobilenet0_conv11"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv12"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv11"
top: "ssd0_mobilenet0_conv12"
convolution_param {
num_output: 128
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm12/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv12"
top: "ssd0_mobilenet0_conv12"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv13"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv12"
top: "ssd0_mobilenet0_conv13"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 3
group: 128
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm13/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv13"
top: "ssd0_mobilenet0_conv13"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv14"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv13"
top: "ssd0_mobilenet0_conv14"
convolution_param {
num_output: 128
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm14/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv14"
top: "ssd0_mobilenet0_conv14"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv15"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv14"
top: "ssd0_mobilenet0_conv15"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 3
group: 128
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm15/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv15"
top: "ssd0_mobilenet0_conv15"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv16"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv15"
top: "ssd0_mobilenet0_conv16"
convolution_param {
num_output: 128
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm16/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv16"
top: "ssd0_mobilenet0_conv16"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv17"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv16"
top: "ssd0_mobilenet0_conv17"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 3
group: 128
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm17/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv17"
top: "ssd0_mobilenet0_conv17"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv18"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv17"
top: "ssd0_mobilenet0_conv18"
convolution_param {
num_output: 128
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm18/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv18"
top: "ssd0_mobilenet0_conv18"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv19"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv18"
top: "ssd0_mobilenet0_conv19"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 3
group: 128
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm19/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv19"
top: "ssd0_mobilenet0_conv19"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv20"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv19"
top: "ssd0_mobilenet0_conv20"
convolution_param {
num_output: 128
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm20/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv20"
top: "ssd0_mobilenet0_conv20"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv21"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv20"
top: "ssd0_mobilenet0_conv21"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 3
group: 128
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm21/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv21"
top: "ssd0_mobilenet0_conv21"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv22"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv21"
top: "ssd0_mobilenet0_conv22"
convolution_param {
num_output: 128
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm22/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv22"
top: "ssd0_mobilenet0_conv22"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_convpredictor1_conv0"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv22"
top: "ssd0_convpredictor1_conv0"
convolution_param {
num_output: 12
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose6"
type: "Permute"
bottom: "ssd0_convpredictor1_conv0"
top: "ssd0_transpose6"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten6"
type: "Flatten"
bottom: "ssd0_transpose6"
top: "ssd0_flatten6"
}
layer {
name: "ssd0_mobilenet0_conv23"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv22"
top: "ssd0_mobilenet0_conv23"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 3
group: 128
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm23/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv23"
top: "ssd0_mobilenet0_conv23"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv24"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv23"
top: "ssd0_mobilenet0_conv24"
convolution_param {
num_output: 256
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm24/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv24"
top: "ssd0_mobilenet0_conv24"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv25"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv24"
top: "ssd0_mobilenet0_conv25"
convolution_param {
num_output: 256
bias_term: true
pad: 1
kernel_size: 3
group: 256
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm25/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv25"
top: "ssd0_mobilenet0_conv25"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_mobilenet0_conv26"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv25"
top: "ssd0_mobilenet0_conv26"
convolution_param {
num_output: 256
bias_term: true
pad: 0
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_mobilenet0_batchnorm26/relu"
type: "ReLU"
bottom: "ssd0_mobilenet0_conv26"
top: "ssd0_mobilenet0_conv26"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_convpredictor3_conv0"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv26"
top: "ssd0_convpredictor3_conv0"
convolution_param {
num_output: 16
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose7"
type: "Permute"
bottom: "ssd0_convpredictor3_conv0"
top: "ssd0_transpose7"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten7"
type: "Flatten"
bottom: "ssd0_transpose7"
top: "ssd0_flatten7"
}
layer {
name: "ssd0_expand_trans_conv0"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv26"
top: "ssd0_expand_trans_conv0"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_expand_trans_bn0/relu"
type: "ReLU"
bottom: "ssd0_expand_trans_conv0"
top: "ssd0_expand_trans_conv0"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_expand_conv0"
type: "Convolution"
bottom: "ssd0_expand_trans_conv0"
top: "ssd0_expand_conv0"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_expand_bn0/relu"
type: "ReLU"
bottom: "ssd0_expand_conv0"
top: "ssd0_expand_conv0"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_convpredictor5_conv0"
type: "Convolution"
bottom: "ssd0_expand_conv0"
top: "ssd0_convpredictor5_conv0"
convolution_param {
num_output: 16
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose8"
type: "Permute"
bottom: "ssd0_convpredictor5_conv0"
top: "ssd0_transpose8"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten8"
type: "Flatten"
bottom: "ssd0_transpose8"
top: "ssd0_flatten8"
}
layer {
name: "ssd0_expand_trans_conv1"
type: "Convolution"
bottom: "ssd0_expand_conv0"
top: "ssd0_expand_trans_conv1"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_expand_trans_bn1/relu"
type: "ReLU"
bottom: "ssd0_expand_trans_conv1"
top: "ssd0_expand_trans_conv1"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_expand_conv1"
type: "Convolution"
bottom: "ssd0_expand_trans_conv1"
top: "ssd0_expand_conv1"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_expand_bn1/relu"
type: "ReLU"
bottom: "ssd0_expand_conv1"
top: "ssd0_expand_conv1"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_convpredictor7_conv0"
type: "Convolution"
bottom: "ssd0_expand_conv1"
top: "ssd0_convpredictor7_conv0"
convolution_param {
num_output: 16
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose9"
type: "Permute"
bottom: "ssd0_convpredictor7_conv0"
top: "ssd0_transpose9"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten9"
type: "Flatten"
bottom: "ssd0_transpose9"
top: "ssd0_flatten9"
}
layer {
name: "ssd0_expand_trans_conv2"
type: "Convolution"
bottom: "ssd0_expand_conv1"
top: "ssd0_expand_trans_conv2"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_expand_trans_bn2/relu"
type: "ReLU"
bottom: "ssd0_expand_trans_conv2"
top: "ssd0_expand_trans_conv2"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_expand_conv2"
type: "Convolution"
bottom: "ssd0_expand_trans_conv2"
top: "ssd0_expand_conv2"
convolution_param {
num_output: 64
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_expand_bn2/relu"
type: "ReLU"
bottom: "ssd0_expand_conv2"
top: "ssd0_expand_conv2"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_convpredictor9_conv0"
type: "Convolution"
bottom: "ssd0_expand_conv2"
top: "ssd0_convpredictor9_conv0"
convolution_param {
num_output: 12
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose10"
type: "Permute"
bottom: "ssd0_convpredictor9_conv0"
top: "ssd0_transpose10"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten10"
type: "Flatten"
bottom: "ssd0_transpose10"
top: "ssd0_flatten10"
}
layer {
name: "ssd0_expand_trans_conv3"
type: "Convolution"
bottom: "ssd0_expand_conv2"
top: "ssd0_expand_trans_conv3"
convolution_param {
num_output: 128
bias_term: true
pad: 1
kernel_size: 1
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_expand_trans_bn3/relu"
type: "ReLU"
bottom: "ssd0_expand_trans_conv3"
top: "ssd0_expand_trans_conv3"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_expand_conv3"
type: "Convolution"
bottom: "ssd0_expand_trans_conv3"
top: "ssd0_expand_conv3"
convolution_param {
num_output: 64
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 2
dilation: 1
}
}
layer {
name: "ssd0_expand_bn3/relu"
type: "ReLU"
bottom: "ssd0_expand_conv3"
top: "ssd0_expand_conv3"
relu_param {
negative_slope: 0.0
}
}
layer {
name: "ssd0_convpredictor11_conv0"
type: "Convolution"
bottom: "ssd0_expand_conv3"
top: "ssd0_convpredictor11_conv0"
convolution_param {
num_output: 12
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose11"
type: "Permute"
bottom: "ssd0_convpredictor11_conv0"
top: "ssd0_transpose11"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten11"
type: "Flatten"
bottom: "ssd0_transpose11"
top: "ssd0_flatten11"
}
layer {
name: "ssd0_concat1"
type: "Concat"
bottom: "ssd0_flatten6"
bottom: "ssd0_flatten7"
bottom: "ssd0_flatten8"
bottom: "ssd0_flatten9"
bottom: "ssd0_flatten10"
bottom: "ssd0_flatten11"
top: "ssd0_concat1"
concat_param {
axis: 1
}
}
layer {
name: "ssd0_convpredictor0_conv0"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv22"
top: "ssd0_convpredictor0_conv0"
convolution_param {
num_output: 6
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose0"
type: "Permute"
bottom: "ssd0_convpredictor0_conv0"
top: "ssd0_transpose0"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten0"
type: "Flatten"
bottom: "ssd0_transpose0"
top: "ssd0_flatten0"
}
layer {
name: "ssd0_convpredictor2_conv0"
type: "Convolution"
bottom: "ssd0_mobilenet0_conv26"
top: "ssd0_convpredictor2_conv0"
convolution_param {
num_output: 8
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose1"
type: "Permute"
bottom: "ssd0_convpredictor2_conv0"
top: "ssd0_transpose1"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten1"
type: "Flatten"
bottom: "ssd0_transpose1"
top: "ssd0_flatten1"
}
layer {
name: "ssd0_convpredictor4_conv0"
type: "Convolution"
bottom: "ssd0_expand_conv0"
top: "ssd0_convpredictor4_conv0"
convolution_param {
num_output: 8
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose2"
type: "Permute"
bottom: "ssd0_convpredictor4_conv0"
top: "ssd0_transpose2"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten2"
type: "Flatten"
bottom: "ssd0_transpose2"
top: "ssd0_flatten2"
}
layer {
name: "ssd0_convpredictor6_conv0"
type: "Convolution"
bottom: "ssd0_expand_conv1"
top: "ssd0_convpredictor6_conv0"
convolution_param {
num_output: 8
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose3"
type: "Permute"
bottom: "ssd0_convpredictor6_conv0"
top: "ssd0_transpose3"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten3"
type: "Flatten"
bottom: "ssd0_transpose3"
top: "ssd0_flatten3"
}
layer {
name: "ssd0_convpredictor8_conv0"
type: "Convolution"
bottom: "ssd0_expand_conv2"
top: "ssd0_convpredictor8_conv0"
convolution_param {
num_output: 6
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose4"
type: "Permute"
bottom: "ssd0_convpredictor8_conv0"
top: "ssd0_transpose4"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten4"
type: "Flatten"
bottom: "ssd0_transpose4"
top: "ssd0_flatten4"
}
layer {
name: "ssd0_convpredictor10_conv0"
type: "Convolution"
bottom: "ssd0_expand_conv3"
top: "ssd0_convpredictor10_conv0"
convolution_param {
num_output: 6
bias_term: true
pad: 1
kernel_size: 3
group: 1
stride: 1
dilation: 1
}
}
layer {
name: "ssd0_transpose5"
type: "Permute"
bottom: "ssd0_convpredictor10_conv0"
top: "ssd0_transpose5"
permute_param {
order: 0
order: 2
order: 3
order: 1
}
}
layer {
name: "ssd0_flatten5"
type: "Flatten"
bottom: "ssd0_transpose5"
top: "ssd0_flatten5"
}
layer {
name: "ssd0_concat0"
type: "Concat"
bottom: "ssd0_flatten0"
bottom: "ssd0_flatten1"
bottom: "ssd0_flatten2"
bottom: "ssd0_flatten3"
bottom: "ssd0_flatten4"
bottom: "ssd0_flatten5"
top: "ssd0_concat0"
concat_param {
axis: 1
}
}
layer {
name: "ssd0_reshape6"
type: "Reshape"
bottom: "ssd0_concat0"
top: "ssd0_reshape6"
reshape_param {
shape {
dim: 0
dim: -1
dim: 2
}
}
}
layer {
name: "softmax0"
type: "Softmax"
bottom: "ssd0_reshape6"
top: "softmax0"
softmax_param {
axis: 2
}
}
layer {
name: "flatten0"
type: "Flatten"
bottom: "softmax0"
top: "flatten0"
}
layer {
name: "ssd0_mobilenet0_relu22_fwd_priorbox"
type: "PriorBox"
bottom: "ssd0_mobilenet0_conv22"
bottom: "data"
top: "ssd0_mobilenet0_relu22_fwd_priorbox"
prior_box_param {
min_size: 51.20000076293945
max_size: 102.4000015258789
aspect_ratio: 1.0
aspect_ratio: 2.0
flip: false
clip: false
variance: 0.10000000149011612
variance: 0.10000000149011612
variance: 0.20000000298023224
variance: 0.20000000298023224
offset: 0.5
}
}
layer {
name: "ssd0_mobilenet0_relu26_fwd_priorbox"
type: "PriorBox"
bottom: "ssd0_mobilenet0_conv26"
bottom: "data"
top: "ssd0_mobilenet0_relu26_fwd_priorbox"
prior_box_param {
min_size: 102.4000015258789
max_size: 189.39999389648438
aspect_ratio: 1.0
aspect_ratio: 2.0
aspect_ratio: 3.0
flip: false
clip: false
variance: 0.10000000149011612
variance: 0.10000000149011612
variance: 0.20000000298023224
variance: 0.20000000298023224
offset: 0.5
}
}
layer {
name: "ssd0_expand_reu0_priorbox"
type: "PriorBox"
bottom: "ssd0_expand_conv0"
bottom: "data"
top: "ssd0_expand_reu0_priorbox"
prior_box_param {
min_size: 189.39999389648438
max_size: 276.3999938964844
aspect_ratio: 1.0
aspect_ratio: 2.0
aspect_ratio: 3.0
flip: false
clip: false
variance: 0.10000000149011612
variance: 0.10000000149011612
variance: 0.20000000298023224
variance: 0.20000000298023224
offset: 0.5
}
}
layer {
name: "ssd0_expand_reu1_priorbox"
type: "PriorBox"
bottom: "ssd0_expand_conv1"
bottom: "data"
top: "ssd0_expand_reu1_priorbox"
prior_box_param {
min_size: 276.3999938964844
max_size: 363.5199890136719
aspect_ratio: 1.0
aspect_ratio: 2.0
aspect_ratio: 3.0
flip: false
clip: false
variance: 0.10000000149011612
variance: 0.10000000149011612
variance: 0.20000000298023224
variance: 0.20000000298023224
offset: 0.5
}
}
layer {
name: "ssd0_expand_reu2_priorbox"
type: "PriorBox"
bottom: "ssd0_expand_conv2"
bottom: "data"
top: "ssd0_expand_reu2_priorbox"
prior_box_param {
min_size: 363.5199890136719
max_size: 450.6000061035156
aspect_ratio: 1.0
aspect_ratio: 2.0
flip: false
clip: false
variance: 0.10000000149011612
variance: 0.10000000149011612
variance: 0.20000000298023224
variance: 0.20000000298023224
offset: 0.5
}
}
layer {
name: "ssd0_expand_reu3_priorbox"
type: "PriorBox"
bottom: "ssd0_expand_conv3"
bottom: "data"
top: "ssd0_expand_reu3_priorbox"
prior_box_param {
min_size: 450.6000061035156
max_size: 492.0
aspect_ratio: 1.0
aspect_ratio: 2.0
flip: false
clip: false
variance: 0.10000000149011612
variance: 0.10000000149011612
variance: 0.20000000298023224
variance: 0.20000000298023224
offset: 0.5
}
}
layer {
name: "concat0"
type: "Concat"
bottom: "ssd0_mobilenet0_relu22_fwd_priorbox"
bottom: "ssd0_mobilenet0_relu26_fwd_priorbox"
bottom: "ssd0_expand_reu0_priorbox"
bottom: "ssd0_expand_reu1_priorbox"
bottom: "ssd0_expand_reu2_priorbox"
bottom: "ssd0_expand_reu3_priorbox"
top: "concat0"
concat_param {
axis: 2
}
}
layer {
name: "detection_out"
type: "DetectionOutput"
bottom: "ssd0_concat1"
bottom: "flatten0"
bottom: "concat0"
top: "detection_out"
detection_output_param {
num_classes: 2
share_location: true
background_label_id: 0
nms_param {
nms_threshold: 0.44999998807907104
top_k: 400
}
code_type: CENTER_SIZE
keep_top_k: 100
confidence_threshold: 0.009999999776482582
}
}
Loading...
马建仓 AI 助手
尝试更多
代码解读
代码找茬
代码优化
Python
1
https://gitee.com/sunsonzh/Mobilenet-SSD-License-Plate-Detection.git
[email protected]:sunsonzh/Mobilenet-SSD-License-Plate-Detection.git
sunsonzh
Mobilenet-SSD-License-Plate-Detection
Mobilenet-SSD-License-Plate-Detection
master

搜索帮助