Repository: ethz-asl/hierarchical_loc Branch: master Commit: 3c9e32e9e01e Files: 78 Total size: 248.4 MB Directory structure: gitextract_gqab9l_y/ ├── .flake8 ├── .gitattributes ├── .gitignore ├── .gitmodules ├── LICENSE ├── README.md ├── global-loc/ │ ├── CMakeLists.txt │ ├── app/ │ │ ├── build-index.cc │ │ └── time-query.cc │ ├── data/ │ │ ├── images/ │ │ │ └── euroc_sample.pgm │ │ ├── lindenhof_afternoon_aligned_mobilenet-d0.35.pb │ │ ├── lindenhof_afternoon_aligned_netvlad.pb │ │ ├── lindenhof_afternoon_aligned_resnet-pool.pb │ │ ├── lindenhof_wet_aligned_mobilenet-d0.35.pb │ │ ├── lindenhof_wet_aligned_netvlad.pb │ │ └── lindenhof_wet_aligned_resnet-pool.pb │ ├── include/ │ │ └── global-loc/ │ │ ├── kd-tree-index.h │ │ ├── pca-reduction.h │ │ ├── place-retrieval.h │ │ └── tensorflow-net.h │ ├── models/ │ │ └── mobilenetvlad_depth-0.35/ │ │ ├── saved_model.pb │ │ └── variables/ │ │ ├── variables.data-00000-of-00001 │ │ └── variables.index │ ├── package.xml │ ├── proto/ │ │ └── global-loc/ │ │ └── descriptor_index.proto │ ├── src/ │ │ └── place-retrieval.cc │ └── test/ │ ├── test_build_index.cc │ ├── test_inference.cc │ ├── test_opencv.cc │ ├── test_query_index.cc │ └── test_tensorflow.cc ├── notebooks/ │ ├── generate_proto_py.sh │ ├── nclt_evaluation.ipynb │ ├── nclt_generate_poses.ipynb │ ├── nclt_generate_triplets.ipynb │ ├── nclt_visualize_preprocessing.ipynb │ ├── nclt_visualize_retrieval.ipynb │ ├── tango_evaluation.ipynb │ ├── tango_visualize_retrieval.ipynb │ └── utils.py └── retrievalnet/ ├── downloading/ │ └── download_google_landmarks.py ├── makefile ├── requirements.txt ├── retrievalnet/ │ ├── __init__.py │ ├── configs/ │ │ ├── delf_train_triplets.yaml │ │ ├── mobilenetvlad_export_nclt.yaml │ │ ├── mobilenetvlad_train_distill.yaml │ │ ├── netvlad_export_distill.yaml │ │ ├── netvlad_export_nclt.yaml │ │ ├── netvlad_train_triplets.yaml │ │ └── resnet_export_nclt.yaml │ ├── datasets/ │ │ ├── __init__.py │ │ ├── base_dataset.py │ │ ├── descriptor_distillation.py │ │ └── nclt.py │ ├── evaluation.py │ ├── export_descriptors.py │ ├── export_model.py │ ├── models/ │ │ ├── __init__.py │ │ ├── backbones/ │ │ │ ├── mobilenet_v2.py │ │ │ ├── resnet_v1.py │ │ │ └── utils/ │ │ │ ├── __init__.py │ │ │ ├── conv_blocks.py │ │ │ ├── mobilenet.py │ │ │ └── resnet_utils.py │ │ ├── base_model.py │ │ ├── delf.py │ │ ├── delf_triplets.py │ │ ├── layers.py │ │ ├── mobilenetvlad.py │ │ ├── netvlad_original.py │ │ └── netvlad_triplets.py │ ├── train.py │ └── utils/ │ ├── __init__.py │ ├── stdout_capturing.py │ └── tools.py ├── setup.py └── setup.sh ================================================ FILE CONTENTS ================================================ ================================================ FILE: .flake8 ================================================ [flake8] max-line-length=89 ================================================ FILE: .gitattributes ================================================ notebooks/* linguist-documentation ================================================ FILE: .gitignore ================================================ env/ __pycache__/ *.egg-info/ settings.py .ipynb_checkpoints/ ================================================ FILE: .gitmodules ================================================ [submodule "catkin_dependencies/tensorflow_catkin"] path = catkin_dependencies/tensorflow_catkin url = https://github.com/ethz-asl/tensorflow_catkin [submodule "catkin_dependencies/protobuf_catkin"] path = catkin_dependencies/protobuf_catkin url = https://github.com/Skydes/protobuf_catkin branch = tensorflow_compatible [submodule "catkin_dependencies/eigen_catkin"] path = catkin_dependencies/eigen_catkin url = https://github.com/Skydes/eigen_catkin branch = tensorflow-compatible [submodule "catkin_dependencies/maplab"] path = catkin_dependencies/maplab url = https://github.com/ethz-asl/maplab branch = pre_release_public/july-2018 [submodule "catkin_dependencies/maplab_dependencies"] path = catkin_dependencies/maplab_dependencies url = https://github.com/ethz-asl/maplab_dependencies branch = pre_release_public/july-2018 ================================================ FILE: LICENSE ================================================ BSD 3-Clause License Copyright (c) 2018, Autonomous Systems Lab All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: * Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. * Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. * Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. ================================================ FILE: README.md ================================================ # Hierarchical Localization :warning: :warning: **For a clean and research-friendly implementation of Hierarchical Localization, please refer to our CVPR 2019 paper at [ethz-asl/hfnet](https://github.com/ethz-asl/hfnet).** :warning: :warning: This repository contains the training and deployment code used in our paper *[Leveraging Deep Visual Descriptors for Hierarchical Efficient Localization](https://arxiv.org/abs/1809.01019)* presented at [CoRL 2018](http://www.robot-learning.org/). This work introduces **MobileNetVLAD**, a mobile-friendly image retrieval deep neural network that significantly improves the performance of classical 6-DoF visual localization through a hierarchical search.
The approach is described in details in our video (click to play).
Retrieval on our Zurich dataset: strong illumination and viewpoint changes.
D?74,
%.:CNKU^gjhdmi]\VOC9314774-'83.2-&""$
"9S[R@=70/4612.33*
@FJIGGKNPPKSQWgwZ9,& %GlACDG>@>;=?<>DC>GBDDB@?>A>8' #!SשN!VkzufUXgjlniTD1'%+9DLPUNPJC@:7+-& ",Sty|xryyh^V5 !h~겫{}{|~|zsj6
+87GLo}ȵjF0
"(4Jpm*%%$")&(./14.43*+4FC-)$8CDDB=.3a$=BbE7-#
" 'E=.-1,
$*387DKQZ\]Wgj_`]USFA@9:5( ?IKJJKLOPOOOPZpK0(%#!%RcEEEBE@D=@BBBDDCGB=@G@AB@C>'!# ([ΥA%\rttfS]jjbf^R9#%,8GHMF@@801-()#"!
&Eeqvl`[axl8*$
'6G^~}淲|{y{}um;
259FZzst||~~r[C1$
#-$")+%#!(0 ?;FDO\}_UTI>A6.3-+&(
!# **5>I[ox}R& >rg1.)$'" ((-*+*/a¾irɒ1(\ozfS:?GWWYMB,ɪtqrmompqn{ȼbI;2,*Cno@Dis9'4eݺ˹j38ANl½wrtؼׁxqqt[ccW__`oofgkru~x|vrrlqnmouzw|ePS^Wb`arnqqywksnmr{tw}H,'+@uH*$6R||r(C~ӚQ0!())%& HPT_bhf^^kyy~o[QJFA?@FLVYXS\`_M@HJ*
0=5AGEJRWML[Y`b`a_ho{n[QE<@79:FVWMQcc)1;5;@FORPRSX]^^cnvzwudPOE>79BGYYVQF/9>FHRY]_^acfdrvuyx{lfTMOILPURRYCI%+I.6)=F-*:+,!
"
'HG
#"%
!E_jҵrW919YlOBU]**OCCEEHDIVME@J>EJA?G8 )+$&"!" $ !% " &!#$&3Yxx__ky{|}z|{z~{~yy|tvzpmtwssqqnulqkopppqpqnknmff[\nkTLKHF@352/+' #$%&(+/=KQVZ\bZaaa^eebobdb^Q2(**086:67659;:8952) 3:9=<:9;FL䖃{}Ş{{̘cA,$&+43><7/$ vC-492/3C>5/))+.142>GNS\_c`fZmxy|iXLIEED?HJU]V\]b_K>FݮD)/=8>CIITRMSQ\]`^Zcfl|~wZMB<757=J[WRQb`".::ACJONUVV\]bbboq}z{piQMB;;>BOZTVUA,8=BMO\`^fbbfmlw|tofQMOFHKPXP[TA!!H.((@H**:'+!#!
'HF
FSbtϪ鲐nC<@DHKS_uP(6YS@@@C@KXJHLGBEB<>H:%$%(/'"!" !! #!& # #$!Am`dovwvx}}wxxuxvswtmomrupqtolqrpkkeponsofjhme\INm_PJKDD;-!$+)*-+)"""#&)3AS[_c]bc_a`ecbe\B,+*/3676:7989;8<963$
'9=>BB??HKI֦͜z~`C4!&,,300*! lN.44.(-DB3/+)((25?9IQQVb`be^`tzjZJHFEBDDMT`W[ac]R>C߲M'
(>4:CDLTRJMRYc_]Y\gkwz}{zn\OB<637@JXYUScf"4C4BDJLRLU\Z]\\fqwzz|ndZKD:;?AO]ZRQ?1;=GKVY]^fb`gnqxzwvsjeQNJDHMUWS^f@ %36-AD()=!("
&!0FH
#
,X\q~ásuթ`SR\ex[=(6bC>?A>?JNHDNE;FC;)!-25,
'#"#$"!#"$%(Bp`hguruy{~w~xx{ywvquusuuwomrsnqtnnnnpnpoknilrhgecWOBKqUMFJCD:7++&!',0+)*&$ %!6GSX_^gaeadgb_\?.&),856;:;99:8;977."
,2<;@AAFEKJó긬{yȊ\<1)!#+*--0,#
d`
$15+*,>@1+*-*&-18>HMT]]b]cV]oz}iTPNFDDEBMU`[U__`Q?FشL--D5 GKQQQVY_[^aenv||zrgUHD>;9CRY\XV=27:BGT[e_bcclhqv{z}rnaVDGOJRSWW`jC"!=(J7LG)(;$*$
!"#/KD
A\Ysm`imȝ}^d{qeTNJ5(5S8@==@BKMAEKDBB<8?A
('%( # $! ##%2WՍafmpuz{|v}yzutvtvsvtpmzooopnnrmnrnlmjfndjmffinh]TJ>ANePMLFC:98697165'$##--+*(!#*=M[V`^_d`ccedR8$%.,4626:894<:5:;5/
4>=@DDACEIMӷŲȴ㶂vD
#$/.,-%!
\o
'*5)*(=C2.,'''*05=ESS]bdcc_bktmXQFDECBJKQZ_XXb^SDCN2$
*>6>BBKPKLOY]d^_aZilw{nXNC@=89;F[]PRfe$ 4C;>ELMOQY]\Xdeenv~}wmcQMG@<=BR_]UV=188CNVZeadignjqr{xvxocULBAKKUXWcjB$(TOL;NK%$=,#2'
"#&1PG
!W_hۻ°zgZR[a㼕xRDCJMFLIDG6.GS>OJJIIPPGESP@EA9=?
'" % ! %&"!!$' $$##,I{`lpsuqyvtrvusufZWXbkrqrslrqvppqpnniilmjtighgfYOKJ>@[cMPMK?898;?:>=HA61*.# !!%%,)%'$'7AOV\_a\^[ab\N4%&,136;88:5456;<=7.5;?DCDCCGJKÿv
%%*.+..$ R}!/0&"&7G1/,((&+26?FOM[_\\a^_mwzgWNFJEAAHEO\[V[a^S?CܼR-"
0@8=BEHRTHNXY`_\_[gho|t{rTMID<87=I[[PTch/<7BDLPRWY]]X^_ilw|vwynbVJ?>;9@X_`XW7/9=HLV]^cdfchlns{|zqgZOGD@IKRVW^p?&,]Q;
67?11C--Kc>@+%J:3;11$ PZ.>XV?;>SaWE;57=5?P( "(',-8815050( "IeE
..5-$6G@.!
%,7!$!0AKQ4 )'2F+
!!#!,,'0,0..-++-1.+./-.+,-*&%++,..1/,+,,,+-/+2/-14582-7^aYMFD>CLXvĻqYRY^kz̻vjuov{Y1+-+((%,FU0)(')(*),&)-.*,1323359722547;51898=8;77856-1220(()8VM1%"" $'(124- /FO6!"
#4B<, (?RH4 -2)(-89:6:;5*.CC6(
mΰ
%,#$)*4@GGOWaovx|{|zvtqh`[SOKLA@=IEA@A?@.j
$6-2795::;