Haar Wavelet Based Block Autoregressive Flows for Trajectories View Full Text


Ontology type: schema:Chapter      Open Access: True


Chapter Info

DATE

2021-03-17

AUTHORS

Apratim Bhattacharyya , Christoph-Nikolas Straehle , Mario Fritz , Bernt Schiele

ABSTRACT

Prediction of trajectories such as that of pedestrians is crucial to the performance of autonomous agents. While previous works have leveraged conditional generative models like GANs and VAEs for learning the likely future trajectories, accurately modeling the dependency structure of these multimodal distributions, particularly over long time horizons remains challenging. Normalizing flow based generative models can model complex distributions admitting exact inference. These include variants with split coupling invertible transformations that are easier to parallelize compared to their autoregressive counterparts. To this end, we introduce a novel Haar wavelet based block autoregressive model leveraging split couplings, conditioned on coarse trajectories obtained from Haar wavelet based transformations at different levels of granularity. This yields an exact inference method that models trajectories at different spatio-temporal resolutions in a hierarchical manner. We illustrate the advantages of our approach for generating diverse and accurate trajectories on two real-world datasets – Stanford Drone and Intersection Drone. More... »

PAGES

275-288

Identifiers

URI

http://scigraph.springernature.com/pub.10.1007/978-3-030-71278-5_20

DOI

http://dx.doi.org/10.1007/978-3-030-71278-5_20

DIMENSIONS

https://app.dimensions.ai/details/publication/pub.1136443618


Indexing Status Check whether this publication has been indexed by Scopus and Web Of Science using the SN Indexing Status Tool
Incoming Citations Browse incoming citations for this publication using opencitations.net

JSON-LD is the canonical representation for SciGraph data.

TIP: You can open this SciGraph record using an external JSON-LD service: JSON-LD Playground Google SDTT

[
  {
    "@context": "https://springernature.github.io/scigraph/jsonld/sgcontext.json", 
    "about": [
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/08", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Information and Computing Sciences", 
        "type": "DefinedTerm"
      }, 
      {
        "id": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/0801", 
        "inDefinedTermSet": "http://purl.org/au-research/vocabulary/anzsrc-for/2008/", 
        "name": "Artificial Intelligence and Image Processing", 
        "type": "DefinedTerm"
      }
    ], 
    "author": [
      {
        "affiliation": {
          "alternateName": "Max Planck Institute for Informatics, Saarbr\u00fccken, Germany", 
          "id": "http://www.grid.ac/institutes/grid.419528.3", 
          "name": [
            "Max Planck Institute for Informatics, Saarbr\u00fccken, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Bhattacharyya", 
        "givenName": "Apratim", 
        "id": "sg:person.013476651003.32", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013476651003.32"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Bosch Center for Artificial Intelligence, Renningen, Germany", 
          "id": "http://www.grid.ac/institutes/grid.6584.f", 
          "name": [
            "Bosch Center for Artificial Intelligence, Renningen, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Straehle", 
        "givenName": "Christoph-Nikolas", 
        "id": "sg:person.0754317026.12", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0754317026.12"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "CISPA Helmholtz Center for Information Security, Saarbr\u00fccken, Germany", 
          "id": "http://www.grid.ac/institutes/grid.507511.7", 
          "name": [
            "CISPA Helmholtz Center for Information Security, Saarbr\u00fccken, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Fritz", 
        "givenName": "Mario", 
        "id": "sg:person.013361072755.17", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013361072755.17"
        ], 
        "type": "Person"
      }, 
      {
        "affiliation": {
          "alternateName": "Max Planck Institute for Informatics, Saarbr\u00fccken, Germany", 
          "id": "http://www.grid.ac/institutes/grid.419528.3", 
          "name": [
            "Max Planck Institute for Informatics, Saarbr\u00fccken, Germany"
          ], 
          "type": "Organization"
        }, 
        "familyName": "Schiele", 
        "givenName": "Bernt", 
        "id": "sg:person.01174260421.90", 
        "sameAs": [
          "https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01174260421.90"
        ], 
        "type": "Person"
      }
    ], 
    "datePublished": "2021-03-17", 
    "datePublishedReg": "2021-03-17", 
    "description": "Prediction of trajectories such as that of pedestrians is crucial to the performance of autonomous agents. While previous works have leveraged conditional generative models like GANs and VAEs for learning the likely future trajectories, accurately modeling the dependency structure of these multimodal distributions, particularly over long time horizons remains challenging. Normalizing flow based generative models can model complex distributions admitting exact inference. These include variants with split coupling invertible transformations that are easier to parallelize compared to their autoregressive counterparts. To this end, we introduce a novel Haar wavelet based block autoregressive model leveraging split couplings, conditioned on coarse trajectories obtained from Haar wavelet based transformations at different levels of granularity. This yields an exact inference method that models trajectories at different spatio-temporal resolutions in a hierarchical manner. We illustrate the advantages of our approach for generating diverse and accurate trajectories on two real-world datasets \u2013 Stanford Drone and Intersection Drone.", 
    "editor": [
      {
        "familyName": "Akata", 
        "givenName": "Zeynep", 
        "type": "Person"
      }, 
      {
        "familyName": "Geiger", 
        "givenName": "Andreas", 
        "type": "Person"
      }, 
      {
        "familyName": "Sattler", 
        "givenName": "Torsten", 
        "type": "Person"
      }
    ], 
    "genre": "chapter", 
    "id": "sg:pub.10.1007/978-3-030-71278-5_20", 
    "isAccessibleForFree": true, 
    "isPartOf": {
      "isbn": [
        "978-3-030-71277-8", 
        "978-3-030-71278-5"
      ], 
      "name": "Pattern Recognition", 
      "type": "Book"
    }, 
    "keywords": [
      "Haar wavelet", 
      "generative model", 
      "different spatio-temporal resolutions", 
      "conditional generative model", 
      "exact inference method", 
      "prediction of trajectories", 
      "autonomous agents", 
      "hierarchical manner", 
      "accurate trajectories", 
      "autoregressive counterparts", 
      "exact inference", 
      "dependency structure", 
      "inference methods", 
      "wavelets", 
      "drones", 
      "future trajectories", 
      "autoregressive flow", 
      "previous work", 
      "likely future trajectories", 
      "spatio-temporal resolution", 
      "long time horizon", 
      "granularity", 
      "complex distribution", 
      "VAE", 
      "trajectories", 
      "split coupling", 
      "time horizon", 
      "different levels", 
      "pedestrians", 
      "multimodal distributions", 
      "model", 
      "autoregressive model", 
      "invertible transformation", 
      "inference", 
      "performance", 
      "advantages", 
      "prediction", 
      "work", 
      "transformation", 
      "method", 
      "manner", 
      "end", 
      "variants", 
      "resolution", 
      "horizon", 
      "flow", 
      "agents", 
      "structure", 
      "GaN", 
      "distribution", 
      "counterparts", 
      "levels", 
      "coupling", 
      "approach"
    ], 
    "name": "Haar Wavelet Based Block Autoregressive Flows for Trajectories", 
    "pagination": "275-288", 
    "productId": [
      {
        "name": "dimensions_id", 
        "type": "PropertyValue", 
        "value": [
          "pub.1136443618"
        ]
      }, 
      {
        "name": "doi", 
        "type": "PropertyValue", 
        "value": [
          "10.1007/978-3-030-71278-5_20"
        ]
      }
    ], 
    "publisher": {
      "name": "Springer Nature", 
      "type": "Organisation"
    }, 
    "sameAs": [
      "https://doi.org/10.1007/978-3-030-71278-5_20", 
      "https://app.dimensions.ai/details/publication/pub.1136443618"
    ], 
    "sdDataset": "chapters", 
    "sdDatePublished": "2022-11-24T21:14", 
    "sdLicense": "https://scigraph.springernature.com/explorer/license/", 
    "sdPublisher": {
      "name": "Springer Nature - SN SciGraph project", 
      "type": "Organization"
    }, 
    "sdSource": "s3://com-springernature-scigraph/baseset/20221124/entities/gbq_results/chapter/chapter_241.jsonl", 
    "type": "Chapter", 
    "url": "https://doi.org/10.1007/978-3-030-71278-5_20"
  }
]
 

Download the RDF metadata as:  json-ld nt turtle xml License info

HOW TO GET THIS DATA PROGRAMMATICALLY:

JSON-LD is a popular format for linked data which is fully compatible with JSON.

curl -H 'Accept: application/ld+json' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-71278-5_20'

N-Triples is a line-based linked data format ideal for batch operations.

curl -H 'Accept: application/n-triples' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-71278-5_20'

Turtle is a human-readable linked data format.

curl -H 'Accept: text/turtle' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-71278-5_20'

RDF/XML is a standard XML format for linked data.

curl -H 'Accept: application/rdf+xml' 'https://scigraph.springernature.com/pub.10.1007/978-3-030-71278-5_20'


 

This table displays all metadata directly associated to this object as RDF triples.

150 TRIPLES      22 PREDICATES      78 URIs      71 LITERALS      7 BLANK NODES

Subject Predicate Object
1 sg:pub.10.1007/978-3-030-71278-5_20 schema:about anzsrc-for:08
2 anzsrc-for:0801
3 schema:author N7f2f036111dc480895f6bfc757916f36
4 schema:datePublished 2021-03-17
5 schema:datePublishedReg 2021-03-17
6 schema:description Prediction of trajectories such as that of pedestrians is crucial to the performance of autonomous agents. While previous works have leveraged conditional generative models like GANs and VAEs for learning the likely future trajectories, accurately modeling the dependency structure of these multimodal distributions, particularly over long time horizons remains challenging. Normalizing flow based generative models can model complex distributions admitting exact inference. These include variants with split coupling invertible transformations that are easier to parallelize compared to their autoregressive counterparts. To this end, we introduce a novel Haar wavelet based block autoregressive model leveraging split couplings, conditioned on coarse trajectories obtained from Haar wavelet based transformations at different levels of granularity. This yields an exact inference method that models trajectories at different spatio-temporal resolutions in a hierarchical manner. We illustrate the advantages of our approach for generating diverse and accurate trajectories on two real-world datasets – Stanford Drone and Intersection Drone.
7 schema:editor N9e13594902ce4c948863364e66438db5
8 schema:genre chapter
9 schema:isAccessibleForFree true
10 schema:isPartOf N08da3c55c8ae451193bcefcb7e045544
11 schema:keywords GaN
12 Haar wavelet
13 VAE
14 accurate trajectories
15 advantages
16 agents
17 approach
18 autonomous agents
19 autoregressive counterparts
20 autoregressive flow
21 autoregressive model
22 complex distribution
23 conditional generative model
24 counterparts
25 coupling
26 dependency structure
27 different levels
28 different spatio-temporal resolutions
29 distribution
30 drones
31 end
32 exact inference
33 exact inference method
34 flow
35 future trajectories
36 generative model
37 granularity
38 hierarchical manner
39 horizon
40 inference
41 inference methods
42 invertible transformation
43 levels
44 likely future trajectories
45 long time horizon
46 manner
47 method
48 model
49 multimodal distributions
50 pedestrians
51 performance
52 prediction
53 prediction of trajectories
54 previous work
55 resolution
56 spatio-temporal resolution
57 split coupling
58 structure
59 time horizon
60 trajectories
61 transformation
62 variants
63 wavelets
64 work
65 schema:name Haar Wavelet Based Block Autoregressive Flows for Trajectories
66 schema:pagination 275-288
67 schema:productId N5dd5e382efde4ffa8b41a76f9e7f6ba6
68 Nc9c4919a537a4498b39cec93dc2976f6
69 schema:publisher N2a2fe89d22f34be7963d4a12e1f19450
70 schema:sameAs https://app.dimensions.ai/details/publication/pub.1136443618
71 https://doi.org/10.1007/978-3-030-71278-5_20
72 schema:sdDatePublished 2022-11-24T21:14
73 schema:sdLicense https://scigraph.springernature.com/explorer/license/
74 schema:sdPublisher Ne80f989d6cb7486a8dd1c6a065276558
75 schema:url https://doi.org/10.1007/978-3-030-71278-5_20
76 sgo:license sg:explorer/license/
77 sgo:sdDataset chapters
78 rdf:type schema:Chapter
79 N076200b3594445cfa3da9e3681626779 rdf:first Nfa4bbf7b3c094130bda7d77ec7e17c53
80 rdf:rest Neee3e0621b1449cb91bfb759faa74ab4
81 N08da3c55c8ae451193bcefcb7e045544 schema:isbn 978-3-030-71277-8
82 978-3-030-71278-5
83 schema:name Pattern Recognition
84 rdf:type schema:Book
85 N0d3b20831c96477bb1a04ff7dbc6ac88 rdf:first sg:person.0754317026.12
86 rdf:rest Nce84d7022830436ea9c44dca113bbe5a
87 N27ed519daeab44ceb66ffe124b3dfc04 schema:familyName Akata
88 schema:givenName Zeynep
89 rdf:type schema:Person
90 N2a2fe89d22f34be7963d4a12e1f19450 schema:name Springer Nature
91 rdf:type schema:Organisation
92 N5dd5e382efde4ffa8b41a76f9e7f6ba6 schema:name dimensions_id
93 schema:value pub.1136443618
94 rdf:type schema:PropertyValue
95 N7f2f036111dc480895f6bfc757916f36 rdf:first sg:person.013476651003.32
96 rdf:rest N0d3b20831c96477bb1a04ff7dbc6ac88
97 N918b5154f70a4a7986d7ae89f97e654b rdf:first sg:person.01174260421.90
98 rdf:rest rdf:nil
99 N9e13594902ce4c948863364e66438db5 rdf:first N27ed519daeab44ceb66ffe124b3dfc04
100 rdf:rest N076200b3594445cfa3da9e3681626779
101 Nab786233a9e041a2952d4d8fe848beb9 schema:familyName Sattler
102 schema:givenName Torsten
103 rdf:type schema:Person
104 Nc9c4919a537a4498b39cec93dc2976f6 schema:name doi
105 schema:value 10.1007/978-3-030-71278-5_20
106 rdf:type schema:PropertyValue
107 Nce84d7022830436ea9c44dca113bbe5a rdf:first sg:person.013361072755.17
108 rdf:rest N918b5154f70a4a7986d7ae89f97e654b
109 Ne80f989d6cb7486a8dd1c6a065276558 schema:name Springer Nature - SN SciGraph project
110 rdf:type schema:Organization
111 Neee3e0621b1449cb91bfb759faa74ab4 rdf:first Nab786233a9e041a2952d4d8fe848beb9
112 rdf:rest rdf:nil
113 Nfa4bbf7b3c094130bda7d77ec7e17c53 schema:familyName Geiger
114 schema:givenName Andreas
115 rdf:type schema:Person
116 anzsrc-for:08 schema:inDefinedTermSet anzsrc-for:
117 schema:name Information and Computing Sciences
118 rdf:type schema:DefinedTerm
119 anzsrc-for:0801 schema:inDefinedTermSet anzsrc-for:
120 schema:name Artificial Intelligence and Image Processing
121 rdf:type schema:DefinedTerm
122 sg:person.01174260421.90 schema:affiliation grid-institutes:grid.419528.3
123 schema:familyName Schiele
124 schema:givenName Bernt
125 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.01174260421.90
126 rdf:type schema:Person
127 sg:person.013361072755.17 schema:affiliation grid-institutes:grid.507511.7
128 schema:familyName Fritz
129 schema:givenName Mario
130 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013361072755.17
131 rdf:type schema:Person
132 sg:person.013476651003.32 schema:affiliation grid-institutes:grid.419528.3
133 schema:familyName Bhattacharyya
134 schema:givenName Apratim
135 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.013476651003.32
136 rdf:type schema:Person
137 sg:person.0754317026.12 schema:affiliation grid-institutes:grid.6584.f
138 schema:familyName Straehle
139 schema:givenName Christoph-Nikolas
140 schema:sameAs https://app.dimensions.ai/discover/publication?and_facet_researcher=ur.0754317026.12
141 rdf:type schema:Person
142 grid-institutes:grid.419528.3 schema:alternateName Max Planck Institute for Informatics, Saarbrücken, Germany
143 schema:name Max Planck Institute for Informatics, Saarbrücken, Germany
144 rdf:type schema:Organization
145 grid-institutes:grid.507511.7 schema:alternateName CISPA Helmholtz Center for Information Security, Saarbrücken, Germany
146 schema:name CISPA Helmholtz Center for Information Security, Saarbrücken, Germany
147 rdf:type schema:Organization
148 grid-institutes:grid.6584.f schema:alternateName Bosch Center for Artificial Intelligence, Renningen, Germany
149 schema:name Bosch Center for Artificial Intelligence, Renningen, Germany
150 rdf:type schema:Organization
 




Preview window. Press ESC to close (or click here)


...