Filters








55,523 Hits in 4.1 sec

Generalized Continual Zero-Shot Learning [article]

Chandan Gautam, Sethupathy Parameswaran, Ashish Mishra, Suresh Sundaram
<span title="2021-02-01">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Recently, zero-shot learning (ZSL) emerged as an exciting topic and attracted a lot of attention.  ...  We develop baselines and evaluate generalized CZSL on five ZSL benchmark datasets for two different settings of continual learning: with and without class incremental.  ...  Generalized Continual Zero-Shot Learning In this section, the continual learning method is proposed for the ZSL framework.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2011.08508v3">arXiv:2011.08508v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/spue4cbhgjdfdflkwdhpobromi">fatcat:spue4cbhgjdfdflkwdhpobromi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210203192029/https://arxiv.org/pdf/2011.08508v3.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/e3/32/e332542f28932fa4104f2a82a8f8f210b4cce4ec.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2011.08508v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Generative Replay-based Continual Zero-Shot Learning [article]

Chandan Gautam, Sethupathy Parameswaran, Ashish Mishra, Suresh Sundaram
<span title="2021-06-07">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Zero-shot learning is a new paradigm to classify objects from classes that are not available at training time.  ...  Zero-shot learning (ZSL) methods have attracted considerable attention in recent years because of their ability to classify unseen/novel class examples.  ...  Related Work This section contains related work in three parts: (i) zero-shot learning (ii) continual learning, and (iii) continual zero-shot learning Zero-shot Learning ZSL is initially introduced in  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2101.08894v2">arXiv:2101.08894v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/yfeyhpapdndmrglfso3qhrpkju">fatcat:yfeyhpapdndmrglfso3qhrpkju</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210609201330/https://arxiv.org/pdf/2101.08894v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/09/72/0972c1a53d10924a5919412c80b909a00ba5f0b9.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2101.08894v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Class Normalization for (Continual)? Generalized Zero-Shot Learning [article]

Ivan Skorokhodov, Mohamed Elhoseiny
<span title="2021-04-14">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
However, in the zero-shot learning (ZSL) world, these ideas have received only marginal attention. This work studies normalization in ZSL scenario from both theoretical and practical perspectives.  ...  Finally, we generalize ZSL to a broader problem -- continual ZSL, and introduce some principled metrics and rigorous baselines for this new setup.  ...  Next, we generalize zero-shot learning into a broader setting of continual zero-shot learning. We propose several metrics for it and test our ideas in this new scenario.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2006.11328v2">arXiv:2006.11328v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/7uei6plgcndxjcx3afklbhqtw4">fatcat:7uei6plgcndxjcx3afklbhqtw4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200626053303/https://arxiv.org/pdf/2006.11328v1.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/04/32/04329047773ee868ee368025f75554e35a0eec04.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2006.11328v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Meta-Learned Attribute Self-Gating for Continual Generalized Zero-Shot Learning [article]

Vinay Kumar Verma, Kevin Liang, Nikhil Mehta, Lawrence Carin
<span title="2021-02-23">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We demonstrate this by performing experiments on five standard ZSL datasets (CUB, aPY, AWA1, AWA2 and SUN) in both generalized zero-shot learning and generalized continual zero-shot learning settings.  ...  We propose a meta-continual zero-shot learning (MCZSL) approach to address both these issues.  ...  to as generalized zero-shot learning (GZSL).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2102.11856v1">arXiv:2102.11856v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/cnmrgtvolvb3dedwsa5nvqwu3e">fatcat:cnmrgtvolvb3dedwsa5nvqwu3e</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210227103357/https://arxiv.org/pdf/2102.11856v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/1a/18/1a1871e53f4d7638539cd326c5d5ef805e462399.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2102.11856v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Adversarial Training of Variational Auto-encoders for Continual Zero-shot Learning(A-CZSL) [article]

Subhankar Ghosh
<span title="2021-04-19">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We show our method is superior in class sequentially learning with ZSL(Zero-Shot Learning) and GZSL(Generalized Zero-Shot Learning).  ...  We propose a continual zero-shot learning model(A-CZSL) that is more suitable in real-case scenarios to address the issue that can learn sequentially and distinguish classes the model has not seen during  ...  Continual Zero-shot Learning Reference [7] proposed a continual zero-shot learning model called Generalized continual zero-shot learning(GCZSL), a single head CZSL where the task identity is revealed  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2102.03778v2">arXiv:2102.03778v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/tmosgdlgtbb4hffxrg5vcd2jgy">fatcat:tmosgdlgtbb4hffxrg5vcd2jgy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210422013829/https://arxiv.org/pdf/2102.03778v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/06/44/0644510c1fc9309579699230e9d15eab1cc0707d.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2102.03778v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Domain-Aware Continual Zero-Shot Learning [article]

Kai Yi, Mohamed Elhoseiny
<span title="2021-12-24">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We introduce Domain Aware Continual Zero-Shot Learning (DACZSL), the task of visually recognizing images of unseen categories in unseen domains sequentially.  ...  Our method also learns a class-wise learnable prompt to obtain better class-level text representation, which is used to represent side information to enable zero-shot prediction of future unseen classes  ...  Continual Zero-Shot Learning (DACZSL).  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2112.12989v1">arXiv:2112.12989v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/4nrgoylotvhuhckwwffc3gijqy">fatcat:4nrgoylotvhuhckwwffc3gijqy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220104013723/https://arxiv.org/pdf/2112.12989v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/48/f3/48f3abeb44dd13d5d3a5d6681426032c7ffb35c9.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2112.12989v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

CoLLIE: Continual Learning of Language Grounding from Language-Image Embeddings [article]

Gabriel Skantze, Bram Willemsen
<span title="2021-11-15">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
original zero-shot performance.  ...  Unlike traditional few-shot learning, the model does not just learn new classes and labels, but can also generalize to similar language use.  ...  A APPENDIX Figure 1 : 1 Figure 1: Comparison of CoLLIE to Zero-shot and Few-shot learning. Green boxes show where continual learning is taking place.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2111.07993v1">arXiv:2111.07993v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/mrehauqgj5azfllpev5mn4uehy">fatcat:mrehauqgj5azfllpev5mn4uehy</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211127035249/https://arxiv.org/pdf/2111.07993v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/e6/7f/e67f05de0cd0b7eac8dd47577c8dd5ef1a6d5eec.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2111.07993v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Zero-Shot Learning by Convex Combination of Semantic Embeddings [article]

Mohammad Norouzi and Tomas Mikolov and Samy Bengio and Yoram Singer and Jonathon Shlens and Andrea Frome and Greg S. Corrado and Jeffrey Dean
<span title="2014-03-21">2014</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Proponents of these image embedding systems have stressed their advantages over the traditional classification framing of image understanding, particularly in terms of the promise for zero-shot learning  ...  We show that this simple and direct method confers many of the advantages associated with more complex image embedding schemes, and indeed outperforms state of the art methods on the ImageNet zero-shot  ...  A key component of zero-shot learning is the way a continuous semantic space of class label embeddings is defined.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1312.5650v3">arXiv:1312.5650v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/z5ir4khulnhb3kf6gbewjpobmi">fatcat:z5ir4khulnhb3kf6gbewjpobmi</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200915191404/https://arxiv.org/pdf/1312.5650v2.pdf" title="fulltext PDF download [not primary version]" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <span style="color: #f43e3e;">&#10033;</span> <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/07/54/075464b2d60e1c90fea5d379adaaf613f6ab9266.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/1312.5650v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Continuous Coordination As a Realistic Scenario for Lifelong Learning [article]

Hadi Nekoei, Akilesh Badrinaaraayanan, Aaron Courville, Sarath Chandar
<span title="2021-06-14">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
In this work, we introduce a multi-agent lifelong learning testbed that supports both zero-shot and few-shot settings.  ...  Current deep reinforcement learning (RL) algorithms are still highly task-specific and lack the ability to generalize to new environments.  ...  Our objective is to design a training paradigm that can learn zero-shot and few-shot coordination with unseen agents.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2103.03216v2">arXiv:2103.03216v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/czld45ilrjfaxldxzmzcndfy5e">fatcat:czld45ilrjfaxldxzmzcndfy5e</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210616035633/https://arxiv.org/pdf/2103.03216v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/16/bc/16bc9c08dbbca2f4f0567b2f5e1de8ee2fc1b72a.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2103.03216v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Bookworm continual learning: beyond zero-shot learning and continual learning [article]

Kai Wang, Luis Herranz, Anjan Dutta, Joost van de Weijer
<span title="2020-08-20">2020</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Thus BCL generalizes both continual learning (CL) and zero-shot learning (ZSL).  ...  We observe that conditioning the feature generator on attributes can actually harm the continual learning ability, and propose two variants (joint class-attribute conditioning and asymmetric generation  ...  Figure 1 . 1 Generalized continual learning: (a) continual learning, (b) zero-shot learning, and (c) bookworm continual learning. 1 Computer Vision Center, Autonomous University of Barcelona (UAB) 2  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2006.15176v3">arXiv:2006.15176v3</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/4xpvuzkrwffvzoivry3w7mdtem">fatcat:4xpvuzkrwffvzoivry3w7mdtem</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20200822012154/https://arxiv.org/pdf/2006.15176v3.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2006.15176v3" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Rectification-based Knowledge Retention for Continual Learning [article]

Pravendra Singh, Pratik Mazumder, Piyush Rai, Vinay P. Namboodiri
<span title="2021-03-30">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Our approach can be used in both the zero-shot and non zero-shot task incremental learning settings.  ...  The task incremental learning problem becomes even more challenging when the test set contains classes that are not part of the train set, i.e., a task incremental generalized zero-shot learning problem  ...  Task Incremental Generalized Zero-Shot Learning The task incremental generalized zero-shot learning setting also involves training the model on a sequence of tasks, but each task contains a set of seen  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2103.16597v1">arXiv:2103.16597v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/fi56ub3kaffihpg2apixsvzmg4">fatcat:fi56ub3kaffihpg2apixsvzmg4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20210404113416/https://arxiv.org/pdf/2103.16597v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/7e/48/7e4822e93b68b3432a32f5eba95d02dd3027195f.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2103.16597v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

CLIP-Adapter: Better Vision-Language Models with Feature Adapters [article]

Peng Gao, Shijie Geng, Renrui Zhang, Teli Ma, Rongyao Fang, Yongfeng Zhang, Hongsheng Li, Yu Qiao
<span title="2021-10-09">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
To avoid non-trivial prompt engineering, context optimization has been proposed to learn continuous vectors as task-specific prompts with few-shot training examples.  ...  On downstream tasks, a carefully chosen text prompt is employed to make zero-shot predictions.  ...  Note that when α equals to 0, it is equivalent to Zero-shot CLIP since no new knowledge is learned.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2110.04544v1">arXiv:2110.04544v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/4zzqfhgcqncpjkwamgyab3s4cm">fatcat:4zzqfhgcqncpjkwamgyab3s4cm</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211015222732/https://arxiv.org/pdf/2110.04544v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/c0/40/c04067f03fba2df0c14ea51a170f213eb2983708.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2110.04544v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Continual-T0: Progressively Instructing 50+ Tasks to Language Models Without Forgetting [article]

Thomas Scialom and Tuhin Chakrabarty and Smaranda Muresan
<span title="2022-05-24">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
Language models trained on these instructions show strong zero-shot performance on several standard datasets.  ...  In spite of the limited success of Continual Learning we show that Language Models can be continual learners.  ...  Zero-shot Instruction Combinations Our CT0 model has learned effectively to process different instructions in specific contexts: word level constraint in the context of headline generation, or an emotional  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2205.12393v1">arXiv:2205.12393v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/bafrscx74zh2xbfgg2hrej2qoe">fatcat:bafrscx74zh2xbfgg2hrej2qoe</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220527034331/https://arxiv.org/pdf/2205.12393v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/88/2e/882ebe805c4d87929dbe4d930a1d52e0882d1d3e.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2205.12393v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Zero-Shot Recommender Systems [article]

Hao Ding, Yifei Ma, Anoop Deoras, Yuyang Wang, Hao Wang
<span title="2021-10-12">2021</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
On the other hand, zero-shot learning promises some degree of generalization from an old dataset to an entirely new dataset. In this paper, we explore the possibility of zero-shot learning in RS.  ...  We develop an algorithm, dubbed ZEro-Shot Recommenders (ZESRec), that is trained on an old dataset and generalize to a new one where there are neither overlapping users nor overlapping items, a setting  ...  On the other hand, zero-shot learning promises some degree of generalization from an old dataset to an entirely new dataset. In this paper, we explore the possibility of zero-shot learning in RecSys.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2105.08318v2">arXiv:2105.08318v2</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/tbpnyiaefrhnlnlwsjqj3qlmbu">fatcat:tbpnyiaefrhnlnlwsjqj3qlmbu</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20211016153456/https://arxiv.org/pdf/2105.08318v2.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/87/27/872719ef003ff5f9320ff338b1fbd540b4080051.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2105.08318v2" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>

Prompt-aligned Gradient for Prompt Tuning [article]

Beier Zhu and Yulei Niu and Yucheng Han and Yue Wu and Hanwang Zhang
<span title="2022-05-30">2022</span> <i > arXiv </i> &nbsp; <span class="release-stage" >pre-print</span>
We present Prompt-aligned Gradient, dubbed ProGrad, to prevent prompt tuning from forgetting the the general knowledge learned from VLMs.  ...  Thanks to the large pre-trained vision-language models (VLMs) like CLIP, we can craft a zero-shot classifier by "prompt", e.g., the confidence score of an image being "[CLASS]" can be obtained by using  ...  (a)&(b): Given 1 shot training sample, CoOp's performance severely drops and under-performs zero-shot CLIP by large margins when the training continues.  ... 
<span class="external-identifiers"> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2205.14865v1">arXiv:2205.14865v1</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/hqq2e7jtmrcdzje7gp4cq235h4">fatcat:hqq2e7jtmrcdzje7gp4cq235h4</a> </span>
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220605151921/https://arxiv.org/pdf/2205.14865v1.pdf" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/97/52/97525413c9f4adae4d620cb9ef0067a7cf292400.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener" href="https://arxiv.org/abs/2205.14865v1" title="arxiv.org access"> <button class="ui compact blue labeled icon button serp-button"> <i class="file alternate outline icon"></i> arxiv.org </button> </a>
&laquo; Previous Showing results 1 &mdash; 15 out of 55,523 results