Attributes2Classname: A discriminative model for attribute-based unsupervised zero-shot learning

Download

index.pdf

Date

2017-01-01

Author

Demirel, Berkan
Cinbiş, Ramazan Gökberk
İKİZLER CİNBİŞ, NAZLI

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

41
views

32
downloads

We propose a novel approach for unsupervised zero-shot learning (ZSL) of classes based on their names. Most existing unsupervised ZSL methods aim to learn a model for directly comparing image features and class names. However, this proves to be a difficult task due to dominance of non-visual semantics in underlying vector-space embeddings of class names. To address this issue, we discriminatively learn a word representation such that the similarities between class and combination of attribute names fall in line with the visual similarity. Contrary to the traditional zero-shot learning approaches that are built upon attribute presence, our approach bypasses the laborious attribute-class relation annotations for unseen classes. In addition, our proposed approach renders text-only training possible, hence, the training can be augmented without the need to collect additional image data. The experimental results show that our method yields state-of-the-art results for unsupervised ZSL in three benchmark datasets.

URI

https://hdl.handle.net/11511/56546

DOI

https://doi.org/10.1109/iccv.2017.139

Conference Name

16th IEEE International Conference on Computer Vision (ICCV)

Collections

Department of Computer Engineering, Conference / Seminar

Suggestions

OpenMETU
Core

Student Perceptions and Readiness on Using a Mobile Application for Vocabulary Acquisition Bakay, Şenol; Delialioğlu, Ömer; Savaş, Perihan (null; 2015-04-17) This study investigated English as a Foreign Language students’ readiness to mobile learning and their opinions after the implementation of an instructional approach using m-learning application. A mixed method research design was used to understand the initial characteristics of students and their perceptions on the m-learning instructional approach. 40 students participated in the pre-treatment questionnaire and 25 students participated to the instruction. Quantitative analyses were used to gather informa...
Effectiveness of considering state similarity for reinforcement learning Girgin, Sertan; Polat, Faruk; Alhajj, Reda (2006-01-01) This paper presents a novel approach that locates states with similar sub-policies, and incorporates them into the reinforcement learning framework for better learning performance. This is achieved by identifying common action sequences of states, which are derived from possible optimal policies and reflected into a tree structure. Based on the number of such sequences, we define a similarity function between two states, which helps to reflect updates on the action-value function of a state to all similar s...
Abstraction in reinforcement learning in partially observable environments Çilden, Erkin; Polat, Faruk; Department of Computer Engineering (2014) Reinforcement learning defines a prominent family of unsupervised machine learning methods in autonomous agents perspective. Markov decision process model provides a solid formal basis for reinforcement learning algorithms. Temporal abstraction mechanisms can be built on reinforcement learning and significant performance gain can be achieved. If the full observability assumption of Markov decision process model is relaxed, the resulting model is partially observable Markov decision process, which constitute...
Learning by Automatic Option Discovery from Conditionally Terminating Sequences Girgin, Sertan; Polat, Faruk; Alhajj, Reda (2006-08-28) This paper proposes a novel approach to discover options in the form of conditionally terminating sequences, and shows how they can be integrated into reinforcement learning framework to improve the learning performance. The method utilizes stored histories of possible optimal policies and constructs a specialized tree structure online in order to identify action sequences which are used frequently together with states that are visited during the execution of such sequences. The tree is then used to implici...
Is case-based learning an effective teaching strategy to challenge students' alternative conceptions regarding chemical kinetics? Yalcinkaya, Eylem; TAŞTAN KIRIK, ÖZGECAN; Boz, Yezdan; Yildiran, Demet (Informa UK Limited, 2012-01-01) Background: Case-based learning (CBL) is simply teaching the concept to the students based on the cases. CBL involves a case, which is a scenario based on daily life, and study questions related to the case, which allows students to discuss their ideas. Chemical kinetics is one of the most difficult concepts for students in chemistry. Students have generally low levels of conceptual understanding and many alternative conceptions regarding it.

Citation Formats

B. Demirel, R. G. Cinbiş, and N. İKİZLER CİNBİŞ, “Attributes2Classname: A discriminative model for attribute-based unsupervised zero-shot learning,” presented at the 16th IEEE International Conference on Computer Vision (ICCV), Venice, ITALY, 2017, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/56546.