使用自然语言作为培训视觉识别模型的监督持有巨大的承诺。最近的作品表明,如果在大型训练数据集中的图像和标题之间的对齐形式使用此类监督,则结果对齐模型在零拍摄分类中表现出色,如下游任务2。在本文中,我们专注于挑逗语言监督的哪些部分对于训练零拍摄图像分类模型至关重要。通过广泛和仔细的实验​​,我们表明:1)可以将简单的单词(弓)标题用作数据集中大多数图像标题的替代品。令人惊讶的是,我们观察到这种方法在与单词平衡结合时提高了零拍分类性能。 2)使用船首净化模型,我们可以通过在没有标题的图像上生成伪弓标题来获得更多培训数据。使用真实和伪弓形标题培训的模型达到了更强的零射性能。在ImageNet-1K零拍评估中,我们只使用3M图像标题对的最佳模型,使用15M图像标题对培训的剪辑模型(31.5%VS 31.3%)进行剪辑。
translated by 谷歌翻译
Pre-trained representations are becoming crucial for many NLP and perception tasks. While representation learning in NLP has transitioned to training on raw text without human annotations, visual and vision-language representations still rely heavily on curated training datasets that are expensive or require expert knowledge. For vision applications, representations are mostly learned using datasets with explicit class labels such as Ima-geNet or OpenImages. For vision-language, popular datasets like Conceptual Captions, MSCOCO, or CLIP all involve a non-trivial data collection (and cleaning) process. This costly curation process limits the size of datasets and hence hinders the scaling of trained models. In this paper, we leverage a noisy dataset of over one billion image alt-text pairs, obtained without expensive filtering or post-processing steps in the Conceptual Captions dataset. A simple dual-encoder architecture learns to align visual and language representations of the image and text pairs using a contrastive loss. We show that the scale of our corpus can make up for its noise and leads to state-of-the-art representations even with such a simple learning scheme. Our visual representation achieves strong performance when transferred to classification tasks such as ImageNet and VTAB. The aligned visual and language representations enables zero-shot image classification and also set new state-of-the-art results on Flickr30K and MSCOCO image-text retrieval benchmarks, even when compared with more sophisticated crossattention models. The representations also enable cross-modality search with complex text and text + image queries.
translated by 谷歌翻译
最先进的愿景和愿景和语言模型依靠大规模的Visio-linguisting预借鉴,以获得各种下游任务的良好性能。通常,这种模型通常是跨模态(对比)或多模态(具有早期融合)但不是两者;它们通常只针对特定的方式或任务。有希望的方向将是使用单一整体普遍模型,作为“基础”,目标是一次性的所有方式 - 真正的视觉和语言基础模型应该擅长视力任务,语言任务和交叉和多数模态视觉和语言任务。我们将Flava介绍在这样的模型中,并在跨越这些目标模式的广泛的35个任务上展示令人印象深刻的性能。
translated by 谷歌翻译
Large-scale cross-modal pre-training paradigms have recently shown ubiquitous success on a wide range of downstream tasks, e.g., zero-shot classification, retrieval and image captioning. However, their successes highly rely on the scale and quality of web-crawled data that naturally contain incomplete and noisy information (e.g., wrong or irrelevant content). Existing works either design manual rules to clean data or generate pseudo-targets as auxiliary signals for reducing noise impact, which do not explicitly tackle both the incorrect and incomplete challenges simultaneously. In this paper, to automatically mitigate the impact of noise by solely mining over existing data, we propose a principled Noise-robust Language-Image Pre-training framework (NLIP) to stabilize pre-training via two schemes: noise-harmonization and noise-completion. First, in noise-harmonization scheme, NLIP estimates the noise probability of each pair according to the memorization effect of cross-modal transformers, then adopts noise-adaptive regularization to harmonize the cross-modal alignments with varying degrees. Second, in noise-completion scheme, to enrich the missing object information of text, NLIP injects a concept-conditioned cross-modal decoder to obtain semantic-consistent synthetic captions to complete noisy ones, which uses the retrieved visual concepts (i.e., objects' names) for the corresponding image to guide captioning generation. By collaboratively optimizing noise-harmonization and noise-completion schemes, our NLIP can alleviate the common noise effects during image-text pre-training in a more efficient way. Extensive experiments show the significant performance improvements of our NLIP using only 26M data over existing pre-trained models (e.g., CLIP, FILIP and BLIP) on 12 zero-shot classification datasets, MSCOCO image captioning and zero-shot image-text retrieval tasks.
translated by 谷歌翻译
State-of-the-art computer vision systems are trained to predict a fixed set of predetermined object categories. This restricted form of supervision limits their generality and usability since additional labeled data is needed to specify any other visual concept. Learning directly from raw text about images is a promising alternative which leverages a much broader source of supervision. We demonstrate that the simple pre-training task of predicting which caption goes with which image is an efficient and scalable way to learn SOTA image representations from scratch on a dataset of 400 million (image, text) pairs collected from the internet. After pre-training, natural language is used to reference learned visual concepts (or describe new ones) enabling zero-shot transfer of the model to downstream tasks. We study the performance of this approach by benchmarking on over 30 different existing computer vision datasets, spanning tasks such as OCR, action recognition in videos, geo-localization, and many types of fine-grained object classification. The model transfers non-trivially to most tasks and is often competitive with a fully supervised baseline without the need for any dataset specific training. For instance, we match the accuracy of the original ResNet-50 on ImageNet zero-shot without needing to use any of the 1.28 million training examples it was trained on. We release our code and pre-trained model weights at https://github.com/OpenAI/CLIP.
translated by 谷歌翻译
具有大尺度图像文本对的视觉预训练(VLP)在各个领域都表现出卓越的性能。但是,Internet上的图像文本对共存通常缺乏明确的对齐信息,这对于VLP来说是次优的。建议采用现成的对象检测器来利用其他图像标签信息。但是,对象检测器是耗时的,只能识别预定义的对象类别,从而限制了模型容量。受到观察的启发,即文本包含不完整的细粒图像信息,我们介绍了Ideas,该想法代表通过在线多标签识别VLP来增加文本多样性。想法表明,可以在VLP期间共同优化从文本中提取的图像标签的多标签学习。此外,想法可以在线识别有价值的图像标签,以提供更明确的文本监督。全面的实验表明,想法可以显着提高多个下游数据集上的性能,并具有较小的额外计算成本。
translated by 谷歌翻译
对比训练有素的语言图像模型,例如剪辑,Align和Basic,已经证明了对多种具有挑战性的自然分配变化的前所未有的鲁棒性。由于这些语言图像模型与以前的培训方法有多种不同,因此一个重要的问题是导致稳定性增长的原因。我们通过系统的实验研究回答这个问题。具体而言,我们研究了鲁棒性增长的五个不同可能的原因:(i)训练集大小,(ii)培训分配,(iii)在培训时进行语言监督,(iv)测试时语言监督,以及(v)对比损失函数。我们的实验表明,更多样化的训练分布是稳健性增长的主要原因,其他因素几乎没有稳健性。除了实验结果之外,我们还引入了Imagenet捕获,这是一种来自Flickr的原始文本注释的Imagenet版本,以实现语言图像训练的进一步受控实验。
translated by 谷歌翻译
本文提出了一种对比调整,这是一种简单的方法,采用对比训练来对准图像和文本模型,同时仍然利用他们的预训练。在我们的实证研究中,我们发现,锁定的预训练图像模型与解锁文本模型最佳。我们调用这种对比调整“锁定图像文本调整”(LIT TOONING)的实例,该实例仅教导文本模型,从预先训练的图像模型中读出了良好的表示新任务。亮度调谐模型将零拍摄传输到新视觉任务的能力提高,例如图像分类或检索。建议的亮度调整是广泛适用的;它可以使用三种不同的图像文本数据集可靠地使用多种预训练方法(监督和无监督)和多种架构(Reset,Vision变换器和MLP-MILLER)。利用基于变压器的预训练VIT-G / 14型号,LIT调谐模型在想象网测试集中实现了84.5%的零射频传输精度,并且在充满挑战的分发ObjectNet测试集中实现了81.1%。
translated by 谷歌翻译
我们提出了Clip-Lite,一种通过与文本注释的特征对齐方式进行视觉表示学习的信息有效方法。与先前提出的剪辑模型相比,剪辑液在优化其对比学学习目标期间只需要一个负图像文本样本对。我们通过利用信息有效的较低限制来实现这一点,以最大化两个输入模态之间的相互信息。这允许剪辑Lite培训,在获得比夹子的更好的性能的同时具有显着减少的数据和批量尺寸。我们通过在Coco-Tablions数据集上预先绘制来评估剪贴画并对其他数据集进行测试传输。 Clip-Lite在Pascal VOC分类上获得+ 15.4%的映射绝对增益,并在ImageNet上获得A + 22.1%的前1个精度增益,同时与其他更复杂,文本监督模型相当或优越。 Clip-Lite还优于剪辑图像和文本检索,零拍分类和视觉接地。最后,通过在表示学习期间执行显式图像文本对齐,我们显示Clip-Lite可以利用语言语义来鼓励可以在下游任务中使用的无偏见的视觉表示。
translated by 谷歌翻译
探索大规模预处理的基础模型对计算机视觉具有重大兴趣,因为这些模型可以快速转移到许多下游任务中。本文介绍了对比字幕(COCA),这是一种极简主义的设计,旨在为图像文本编码器编码器基础模型预算与对比度损失和字幕损失,从而从剪辑和诸如simvlm之类的生成方法之类的对比方法中包含模型能力。与所有解码器层都参与编码器输出的标准编码器 - 模块变压器相反,可口可乐省略了解码器层的上半部分的交叉注意,以编码单峰文本表示,并串联到剩余的解码器层,这些解码器与图像编码器相交的解码器层多模式图像文本表示。除了对多模态解码器输出的字幕损失外,我们还应用了单峰图像和文本嵌入之间的对比损失,该输出可以预测文本令牌自动加压。通过共享相同的计算图,可以用最小的开销有效地计算两个培训目标。可口可乐是端到端和从头开始的网络尺度alt-text数据和带注释的图像,通过将所有标签视为文本,无缝地统一自然语言监督以进行表示。从经验上讲,可口可乐通过零拍传输或在广泛的下游任务上进行零摄像转移或最少的特定任务适应,跨越视觉识别(Imagenet,Kinetics-400/600/700,瞬间, ),交叉模式检索(MSCOCO,FLICKR30K,MSR-VTT),多模式理解(VQA,SNLI-VE,NLVR2)和图像字幕(MSCOCO,NOCAPS)。值得注意的是,在Imagenet分类方面,COCA获得了86.3%的TOP-1准确性,带有冷冻编码器和学习的分类头90.6%,以及带有填充编码器的Imagenet上的新最先进的91.0%Top-1 Top-1精度。
translated by 谷歌翻译
对比性语言图像预处理(剪辑)受到广泛关注,因为它的学会表示形式可以很好地转移到各种下游任务上。在剪辑训练期间,Infonce目标旨在使正面图像对齐和分开的负面图像对齐。在本文中,我们在此过程中显示了表示分组的效果:Infonce客观间接通过随机出现的模式内锚将语义相似的表示形式组合在一起。我们引入了原型对比度图像预处理(原始的),以提高其效率并提高其针对模态差距的鲁棒性来增强这种分组。具体而言,原始利润在图像和文本空间之间建立了原型级别的歧视,从而有效传输了更高级别的结构知识。我们进一步提出了典型的背部翻译(PBT),以将表示形式分组与表示形式对齐,从而有效地学习了在较大的模态差距下有意义的表示。 PBT还使我们能够以更丰富的先验知识介绍其他外部教师。 ProtoClip通过在线情节培训策略进行了培训,这可以扩展到无限量的数据。结合上述新颖的设计,我们在概念标题上训练原始设计,并获得了 +5.81%的成像网线性探测改进,并且 +2.01%的Imagenet Zero Zero-shot分类改进。代码可在https://github.com/megvii-research/protoclip上找到。
translated by 谷歌翻译
作为人类已知的最直观的界面之一,自然语言有可能调解许多涉及人类计算机互动的任务,尤其是在音乐信息检索等以应用程序为中心的领域。在这项工作中,我们探索了跨模式学习,以试图在音乐领域弥合音频和语言。为此,我们提出了Muscall,这是音乐对比的音频学习框架。我们的方法由双重编码架构组成,该体系结构了解音乐音频和描述性句子对之间的对齐方式,生成可用于文本到原告和音频到文本检索的多模式嵌入。多亏了这个属性,肌肉几乎可以转移到任何可以作为基于文本检索的任务转移到任何任务。我们的实验表明,我们的方法在检索音频时的性能要比基线要好得多,该音频与文本描述匹配,相反,与音频查询匹配的文本。我们还证明,我们的模型的多模式对齐能力可以成功扩展到零摄像转移方案,用于流派分类和在两个公共数据集上自动标记。
translated by 谷歌翻译
最近的工作表明,自我监督的预训练导致对挑战性视觉识别任务的监督学习改进。剪辑是一种令人兴奋的学习语言监督的新方法,展示了各种基准的有希望的表现。在这项工作中,我们探索自我监督的学习是否可以帮助使用语言监督来进行视觉表现学习。我们介绍了一个用于组合自我监督学习和剪辑预训练的多任务学习框架。在使用视觉变形金刚进行预培训之后,我们在三个不同的设置下彻底评估了代表性质量,并将性能与自我监督学习进行了比较:零拍摄传输,线性分类和端到端的FineTuning。在ImageNet和电池的额外数据集中,我们发现SLIP通过大幅度提高了精度。我们将通过关于不同模型大小,培训计划和预训练预训练数据集的实验进行验证。我们的研究结果表明,滑块享有世界上最好的:性能比自我监督更好(+ 8.1%的线性精度)和语言监督(+ 5.2%的零射精精度)。
translated by 谷歌翻译
在这项工作中,我们提出了一种开放式摄制对象检测方法,该方法基于图像映射对,学会了检测新颖对象类别以及给定的一组已知类别。这是一种两阶段的训练方法,首先使用位置引导的图像捕获匹配技术以弱监督的方式学习新颖和已知类别的类标签,第二个使用已知的类注释专用于对象检测任务的模型。我们表明,一个简单的语言模型比检测新对象的大型上下文化语言模型更适合。此外,我们引入了一种一致性调查技术,以更好地利用图像捕获对信息。我们的方法比较与现有的开放式检测方法相比,同时具有数据效率。源代码可从https://github.com/lmb-freiburg/locov获得。
translated by 谷歌翻译
在本文中,我们研究了如何在视觉和语言(V+L)表示学习中使用蒙版的信号建模。与其独立开发蒙面语言建模(MLM)和蒙面图像建模(MIM),我们建议建立关节蒙面的视觉和语言建模,其中一种模态的掩盖信号是在另一种方式的帮助下重建的。这是由图像文本配对数据的性质和文本传达几乎相同的信息但以不同格式传达的。在另一种模态下进行的一种模式的掩盖信号重建也可以隐式学习语言令牌和图像贴片之间的跨模式对齐。我们对各种V+L任务的实验表明,该建议的方法不仅可以通过使用大量数据来实现最先进的性能,而且还可以通过有限的培训数据的制度优于其他竞争对手。
translated by 谷歌翻译
Multimodal models are becoming increasingly effective, in part due to unified components, such as the Transformer architecture. However, multimodal models still often consist of many task- and modality-specific pieces and training procedures. For example, CLIP (Radford et al., 2021) trains independent text and image towers via a contrastive loss. We explore an additional unification: the use of a pure pixel-based model to perform image, text, and multimodal tasks. Our model is trained with contrastive loss alone, so we call it CLIP-Pixels Only (CLIPPO). CLIPPO uses a single encoder that processes both regular images and text rendered as images. CLIPPO performs image-based tasks such as retrieval and zero-shot image classification almost as well as CLIP, with half the number of parameters and no text-specific tower or embedding. When trained jointly via image-text contrastive learning and next-sentence contrastive learning, CLIPPO can perform well on natural language understanding tasks, without any word-level loss (language modelling or masked language modelling), outperforming pixel-based prior work. Surprisingly, CLIPPO can obtain good accuracy in visual question answering, simply by rendering the question and image together. Finally, we exploit the fact that CLIPPO does not require a tokenizer to show that it can achieve strong performance on multilingual multimodal retrieval without
translated by 谷歌翻译
Most existing text-video retrieval methods focus on cross-modal matching between the visual content of offline videos and textual query sentences. However, in real scenarios, online videos are frequently accompanied by relevant text information such as titles, tags, and even subtitles, which can be utilized to match textual queries. This inspires us to generate associated captions from offline videos to help with existing text-video retrieval methods. To do so, we propose to use the zero-shot video captioner with knowledge of pre-trained web-scale models (e.g., CLIP and GPT-2) to generate captions for offline videos without any training. Given the captions, one question naturally arises: what can auxiliary captions do for text-video retrieval? In this paper, we present a novel framework Cap4Video, which makes use of captions from three aspects: i) Input data: The video and captions can form new video-caption pairs as data augmentation for training. ii) Feature interaction: We perform feature interaction between video and caption to yield enhanced video representations. iii) Output score: The Query-Caption matching branch can be complementary to the original Query-Video matching branch for text-video retrieval. We conduct thorough ablation studies to demonstrate the effectiveness of our method. Without any post-processing, our Cap4Video achieves state-of-the-art performance on MSR-VTT (51.4%), VATEX (66.6%), MSVD (51.8%), and DiDeMo (52.0%).
translated by 谷歌翻译
Image token removal is an efficient augmentation strategy for reducing the cost of computing image features. However, this efficient augmentation strategy has been found to adversely affect the accuracy of CLIP-based training. We hypothesize that removing a large portion of image tokens may improperly discard the semantic content associated with a given text description, thus constituting an incorrect pairing target in CLIP training. To address this issue, we propose an attentive token removal approach for CLIP training, which retains tokens with a high semantic correlation to the text description. The correlation scores are computed in an online fashion using the EMA version of the visual encoder. Our experiments show that the proposed attentive masking approach performs better than the previous method of random token removal for CLIP training. The approach also makes it efficient to apply multiple augmentation views to the image, as well as introducing instance contrastive learning tasks between these views into the CLIP framework. Compared to other CLIP improvements that combine different pre-training targets such as SLIP and MaskCLIP, our method is not only more effective, but also much more efficient. Specifically, using ViT-B and YFCC-15M dataset, our approach achieves $43.9\%$ top-1 accuracy on ImageNet-1K zero-shot classification, as well as $62.7/42.1$ and $38.0/23.2$ I2T/T2I retrieval accuracy on Flickr30K and MS COCO, which are $+1.1\%$, $+5.5/+0.9$, and $+4.4/+1.3$ higher than the SLIP method, while being $2.30\times$ faster. An efficient version of our approach running $1.16\times$ faster than the plain CLIP model achieves significant gains of $+5.3\%$, $+11.3/+8.0$, and $+9.5/+4.9$ on these benchmarks.
translated by 谷歌翻译
远见和语言预测已成为解决多模式下游任务的普遍方法。当前的趋势是朝着更大的模型和预处理数据集迈进。从长远来看,这一计算头急促似乎是不合理的,而是朝着可持续的解决方案迈进,事实上,排除了资源有限的学术实验室。在这项工作中,我们提出了一个称为VICHA的新框架,该框架有效利用输入数据以通过以下方式提高学习,以: ,(c)利用图像级注释,称为视觉概念,使用现有基础模型(例如剪辑)获得,以提高图像编码器的性能。尽管对数据的预估计少了四倍,但我们的VICHA策略在下游任务(例如图像文本检索,VQA,视觉推理,视觉上和视觉接地)上的其他方法优于其他方法。该代码将在此处公开提供:https://github.com/mshukor/vicha
translated by 谷歌翻译
可以代表和描述环境声音的机器具有实际潜力,例如,用于音频标记和标题系统。普遍的学习范式已经依赖于并行音频文本数据,但是,Web上几乎没有可用。我们提出了vip-ant,它在不使用任何并行音频文本数据的情况下诱导\ textbf {a} udio- \ textBF {t} EXT对齐。我们的主要思想是在双模形图像文本表示和双模态图像 - 音频表示之间共享图像模型;图像模态用作枢轴,并将音频和文本连接在三模态嵌入空间中。在没有配对的音频文本数据的困难零拍设置中,我们的模型在ESC50和US8K音频分类任务上展示了最先进的零点性能,甚至超过了披肩标题的领域的监督状态检索(带音频查询)2.2 \%R @ 1。我们进一步调查了最小音频监控的情况,发现,例如,只有几百个监督的音频文本对将零拍音频分类精度提高8 \%US8K。然而,为了匹配人类奇偶校验,我们的经验缩放实验表明我们需要大约2米$ 2 ^ {21} \约2M $监督的音频标题对。我们的工作开辟了新的途径,用于学习音频文本连接,几乎没有并行音频文本数据。
translated by 谷歌翻译