Matches in SemOpenAlex for { <https://semopenalex.org/work/W3042830062> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W3042830062 abstract "Learning visual features from unlabeled image data is an important yet challenging task, which is often achieved by training a model on some annotation-free information. We consider spatial contexts, for which we solve so-called jigsaw puzzles, i.e., each image is cut into grids and then disordered, and the goal is to recover the correct configuration. Existing approaches formulated it as a classification task by defining a fixed mapping from a small subset of configurations to a class set, but these approaches ignore the underlying relationship between different configurations and also limit their application to more complex scenarios. This paper presents a novel approach which applies to jigsaw puzzles with an arbitrary grid size and dimensionality. We provide a fundamental and generalized principle, that weaker cues are easier to be learned in an unsupervised manner and also transfer better. In the context of puzzle recognition, we use an iterative manner which, instead of solving the puzzle all at once, adjusts the order of the patches in each step until convergence. In each step, we combine both unary and binary features on each patch into a cost function judging the correctness of the current configuration. Our approach, by taking similarity between puzzles into consideration, enjoys a more reasonable way of learning visual knowledge. We verify the effectiveness of our approach in two aspects. First, it is able to solve arbitrarily complex puzzles, including high-dimensional puzzles, that prior methods are difficult to handle. Second, it serves as a reliable way of network initialization, which leads to better transfer performance in a few visual recognition tasks including image classification, object detection, and semantic segmentation." @default.
- W3042830062 created "2020-07-23" @default.
- W3042830062 creator A5003527952 @default.
- W3042830062 creator A5012409526 @default.
- W3042830062 creator A5026478150 @default.
- W3042830062 creator A5028045660 @default.
- W3042830062 creator A5060664791 @default.
- W3042830062 creator A5075005282 @default.
- W3042830062 creator A5075290241 @default.
- W3042830062 creator A5091670458 @default.
- W3042830062 date "2018-12-02" @default.
- W3042830062 modified "2023-09-23" @default.
- W3042830062 title "Iterative Reorganization with Weak Spatial Constraints: Solving Arbitrary Jigsaw Puzzles for Unsupervised Representation Learning" @default.
- W3042830062 doi "https://doi.org/10.48550/arxiv.1812.00329" @default.
- W3042830062 hasPublicationYear "2018" @default.
- W3042830062 type Work @default.
- W3042830062 sameAs 3042830062 @default.
- W3042830062 citedByCount "0" @default.
- W3042830062 crossrefType "posted-content" @default.
- W3042830062 hasAuthorship W3042830062A5003527952 @default.
- W3042830062 hasAuthorship W3042830062A5012409526 @default.
- W3042830062 hasAuthorship W3042830062A5026478150 @default.
- W3042830062 hasAuthorship W3042830062A5028045660 @default.
- W3042830062 hasAuthorship W3042830062A5060664791 @default.
- W3042830062 hasAuthorship W3042830062A5075005282 @default.
- W3042830062 hasAuthorship W3042830062A5075290241 @default.
- W3042830062 hasAuthorship W3042830062A5091670458 @default.
- W3042830062 hasBestOaLocation W30428300621 @default.
- W3042830062 hasConcept C111030470 @default.
- W3042830062 hasConcept C11413529 @default.
- W3042830062 hasConcept C114466953 @default.
- W3042830062 hasConcept C114614502 @default.
- W3042830062 hasConcept C119857082 @default.
- W3042830062 hasConcept C134306372 @default.
- W3042830062 hasConcept C14036430 @default.
- W3042830062 hasConcept C145420912 @default.
- W3042830062 hasConcept C151730666 @default.
- W3042830062 hasConcept C154945302 @default.
- W3042830062 hasConcept C177148314 @default.
- W3042830062 hasConcept C17744445 @default.
- W3042830062 hasConcept C187691185 @default.
- W3042830062 hasConcept C199360897 @default.
- W3042830062 hasConcept C199539241 @default.
- W3042830062 hasConcept C2524010 @default.
- W3042830062 hasConcept C2776359362 @default.
- W3042830062 hasConcept C2779343474 @default.
- W3042830062 hasConcept C2779405079 @default.
- W3042830062 hasConcept C33923547 @default.
- W3042830062 hasConcept C41008148 @default.
- W3042830062 hasConcept C55439883 @default.
- W3042830062 hasConcept C78023250 @default.
- W3042830062 hasConcept C78458016 @default.
- W3042830062 hasConcept C86803240 @default.
- W3042830062 hasConcept C94625758 @default.
- W3042830062 hasConceptScore W3042830062C111030470 @default.
- W3042830062 hasConceptScore W3042830062C11413529 @default.
- W3042830062 hasConceptScore W3042830062C114466953 @default.
- W3042830062 hasConceptScore W3042830062C114614502 @default.
- W3042830062 hasConceptScore W3042830062C119857082 @default.
- W3042830062 hasConceptScore W3042830062C134306372 @default.
- W3042830062 hasConceptScore W3042830062C14036430 @default.
- W3042830062 hasConceptScore W3042830062C145420912 @default.
- W3042830062 hasConceptScore W3042830062C151730666 @default.
- W3042830062 hasConceptScore W3042830062C154945302 @default.
- W3042830062 hasConceptScore W3042830062C177148314 @default.
- W3042830062 hasConceptScore W3042830062C17744445 @default.
- W3042830062 hasConceptScore W3042830062C187691185 @default.
- W3042830062 hasConceptScore W3042830062C199360897 @default.
- W3042830062 hasConceptScore W3042830062C199539241 @default.
- W3042830062 hasConceptScore W3042830062C2524010 @default.
- W3042830062 hasConceptScore W3042830062C2776359362 @default.
- W3042830062 hasConceptScore W3042830062C2779343474 @default.
- W3042830062 hasConceptScore W3042830062C2779405079 @default.
- W3042830062 hasConceptScore W3042830062C33923547 @default.
- W3042830062 hasConceptScore W3042830062C41008148 @default.
- W3042830062 hasConceptScore W3042830062C55439883 @default.
- W3042830062 hasConceptScore W3042830062C78023250 @default.
- W3042830062 hasConceptScore W3042830062C78458016 @default.
- W3042830062 hasConceptScore W3042830062C86803240 @default.
- W3042830062 hasConceptScore W3042830062C94625758 @default.
- W3042830062 hasLocation W30428300621 @default.
- W3042830062 hasOpenAccess W3042830062 @default.
- W3042830062 hasPrimaryLocation W30428300621 @default.
- W3042830062 hasRelatedWork W161990344 @default.
- W3042830062 hasRelatedWork W2088334249 @default.
- W3042830062 hasRelatedWork W2105917874 @default.
- W3042830062 hasRelatedWork W2139069334 @default.
- W3042830062 hasRelatedWork W2922457425 @default.
- W3042830062 hasRelatedWork W3164900290 @default.
- W3042830062 hasRelatedWork W3172480809 @default.
- W3042830062 hasRelatedWork W4303856621 @default.
- W3042830062 hasRelatedWork W4376482778 @default.
- W3042830062 hasRelatedWork W586543330 @default.
- W3042830062 isParatext "false" @default.
- W3042830062 isRetracted "false" @default.
- W3042830062 magId "3042830062" @default.
- W3042830062 workType "article" @default.