Abstract:In this work, we discuss evaluating video foundation models in a fair and robust manner. Unlike language or image foundation models, many video foundation models are evaluated with differing parameters (such as sampling rate, number of frames, pretraining steps, etc.), making fair and robust comparisons challenging. Therefore, we present a carefully designed evaluation framework for measuring two core capabilities of video comprehension: appearance and motion understanding. Our findings reveal that existing video foundation models, whether text-supervised like UMT or InternVideo2, or self-supervised like V-JEPA, exhibit limitations in at least one of these capabilities. As an alternative, we introduce TWLV-I, a new video foundation model that constructs robust visual representations for both motion- and appearance-based videos. Based on the average top-1 accuracy of linear probing on five action recognition benchmarks, pretrained only on publicly accessible datasets, our model shows a 4.6%p improvement compared to V-JEPA (ViT-L) and a 7.7%p improvement compared to UMT (ViT-L). Even when compared to much larger models, our model demonstrates a 7.2%p improvement compared to DFN (ViT-H), a 2.7%p improvement compared to V-JEPA~(ViT-H) and a 2.8%p improvement compared to InternVideo2 (ViT-g). We provide embedding vectors obtained by TWLV-I from videos of several commonly used video benchmarks, along with evaluation source code that can directly utilize these embeddings. The code is available on "https://github.com/twelvelabs-io/video-embeddings-evaluation-framework".
Abstract:Transfer learning for GANs successfully improves generation performance under low-shot regimes. However, existing studies show that the pretrained model using a single benchmark dataset is not generalized to various target datasets. More importantly, the pretrained model can be vulnerable to copyright or privacy risks as membership inference attack advances. To resolve both issues, we propose an effective and unbiased data synthesizer, namely Primitives-PS, inspired by the generic characteristics of natural images. Specifically, we utilize 1) the generic statistics on the frequency magnitude spectrum, 2) the elementary shape (i.e., image composition via elementary shapes) for representing the structure information, and 3) the existence of saliency as prior. Since our synthesizer only considers the generic properties of natural images, the single model pretrained on our dataset can be consistently transferred to various target datasets, and even outperforms the previous methods pretrained with the natural images in terms of Fr'echet inception distance. Extensive analysis, ablation study, and evaluations demonstrate that each component of our data synthesizer is effective, and provide insights on the desirable nature of the pretrained model for the transferability of GANs.
Abstract:Every recent image-to-image translation model uses either image-level (i.e. input-output pairs) or set-level (i.e. domain labels) supervision at minimum. However, even the set-level supervision can be a serious bottleneck for data collection in practice. In this paper, we tackle image-to-image translation in a fully unsupervised setting, i.e., neither paired images nor domain labels. To this end, we propose the truly unsupervised image-to-image translation method (TUNIT) that simultaneously learns to separate image domains via an information-theoretic approach and generate corresponding images using the estimated domain labels. Experimental results on various datasets show that the proposed method successfully separates domains and translates images across those domains. In addition, our model outperforms existing set-level supervised methods under a semi-supervised setting, where a subset of domain labels is provided. The source code is available at https://github.com/clovaai/tunit
Abstract:We propose a novel framework for simultaneously generating and manipulating the face images with desired attributes. While the state-of-the-art attribute editing technique has achieved the impressive performance for creating realistic attribute effects, they only address the image editing problem, using the input image as the condition of model. Recently, several studies attempt to tackle both novel face generation and attribute editing problem using a single solution. However, their image quality is still unsatisfactory. Our goal is to develop a single unified model that can simultaneously create and edit high quality face images with desired attributes. A key idea of our work is that we decompose the image into the latent and attribute vector in low dimensional representation, and then utilize the GAN framework for mapping the low dimensional representation to the image. In this way, we can address both the generation and editing problem by learning the generator. For qualitative and quantitative evaluations, the proposed algorithm outperforms recent algorithms addressing the same problem. Also, we show that our model can achieve the competitive performance with the state-of-the-art attribute editing technique in terms of attribute editing quality.