This paper introduces a media service model that exploits artificial intelligence (AI) video generators at the receive end. This proposal deviates from the traditional multimedia ecosystem, completely relying on in-house production, by shifting part of the content creation onto the receiver. We bring a semantic process into the framework, allowing the distribution network to provide service elements that prompt the content generator, rather than distributing encoded data of fully finished programs. The service elements include fine-tailored text descriptions, lightweight image data of some objects, or application programming interfaces, comprehensively referred to as semantic sources, and the user terminal translates the received semantic data into video frames. Empowered by the random nature of generative AI, the users could then experience super-personalized services accordingly. The proposed idea incorporates the situations in which the user receives different service providers' element packages; a sequence of packages over time, or multiple packages at the same time. Given promised in-context coherence and content integrity, the combinatory dynamics will amplify the service diversity, allowing the users to always chance upon new experiences. This work particularly aims at short-form videos and advertisements, which the users would easily feel fatigued by seeing the same frame sequence every time. In those use cases, the content provider's role will be recast as scripting semantic sources, transformed from a thorough producer. Overall, this work explores a new form of media ecosystem facilitated by receiver-embedded generative models, featuring both random content dynamics and enhanced delivery efficiency simultaneously.
Stable diffusion is the mainstay of the text-to-image (T2I) synthesis in the community due to its generation performance and open-source nature. Recently, Stable Diffusion XL (SDXL), the successor of stable diffusion, has received a lot of attention due to its significant performance improvements with a higher resolution of 1024x1024 and a larger model. However, its increased computation cost and model size require higher-end hardware(e.g., bigger VRAM GPU) for end-users, incurring higher costs of operation. To address this problem, in this work, we propose an efficient latent diffusion model for text-to-image synthesis obtained by distilling the knowledge of SDXL. To this end, we first perform an in-depth analysis of the denoising U-Net in SDXL, which is the main bottleneck of the model, and then design a more efficient U-Net based on the analysis. Secondly, we explore how to effectively distill the generation capability of SDXL into an efficient U-Net and eventually identify four essential factors, the core of which is that self-attention is the most important part. With our efficient U-Net and self-attention-based knowledge distillation strategy, we build our efficient T2I models, called KOALA-1B & -700M, while reducing the model size up to 54% and 69% of the original SDXL model. In particular, the KOALA-700M is more than twice as fast as SDXL while still retaining a decent generation quality. We hope that due to its balanced speed-performance tradeoff, our KOALA models can serve as a cost-effective alternative to SDXL in resource-constrained environments.
Self-supervised Video Representation Learning (VRL) aims to learn transferrable representations from uncurated, unlabeled video streams that could be utilized for diverse downstream tasks. With recent advances in Masked Image Modeling (MIM), in which the model learns to predict randomly masked regions in the images given only the visible patches, MIM-based VRL methods have emerged and demonstrated their potential by significantly outperforming previous VRL methods. However, they require an excessive amount of computations due to the added temporal dimension. This is because existing MIM-based VRL methods overlook spatial and temporal inequality of information density among the patches in arriving videos by resorting to random masking strategies, thereby wasting computations on predicting uninformative tokens/frames. To tackle these limitations of Masked Video Modeling, we propose a new token selection method that masks our more important tokens according to the object's motions in an online manner, which we refer to as Motion-centric Token Selection. Further, we present a dynamic frame selection strategy that allows the model to focus on informative and causal frames with minimal redundancy. We validate our method over multiple benchmark and Ego4D datasets, showing that the pre-trained model using our proposed method significantly outperforms state-of-the-art VRL methods on downstream tasks, such as action recognition and object state change classification while largely reducing memory requirements during pre-training and fine-tuning.
Masked image modeling (MIM) has become a popular strategy for self-supervised learning~(SSL) of visual representations with Vision Transformers. A representative MIM model, the masked auto-encoder (MAE), randomly masks a subset of image patches and reconstructs the masked patches given the unmasked patches. Concurrently, many recent works in self-supervised learning utilize the student/teacher paradigm which provides the student with an additional target based on the output of a teacher composed of an exponential moving average (EMA) of previous students. Although common, relatively little is known about the dynamics of the interaction between the student and teacher. Through analysis on a simple linear model, we find that the teacher conditionally removes previous gradient directions based on feature similarities which effectively acts as a conditional momentum regularizer. From this analysis, we present a simple SSL method, the Reconstruction-Consistent Masked Auto-Encoder (RC-MAE) by adding an EMA teacher to MAE. We find that RC-MAE converges faster and requires less memory usage than state-of-the-art self-distillation methods during pre-training, which may provide a way to enhance the practicality of prohibitively expensive self-supervised learning of Vision Transformer models. Additionally, we show that RC-MAE achieves more robustness and better performance compared to MAE on downstream tasks such as ImageNet-1K classification, object detection, and instance segmentation.
Dense computer vision tasks such as object detection and segmentation require effective multi-scale feature representation for detecting or classifying objects or regions with varying sizes. While Convolutional Neural Networks (CNNs) have been the dominant architectures for such tasks, recently introduced Vision Transformers (ViTs) aim to replace them as a backbone. Similar to CNNs, ViTs build a simple multi-stage structure (i.e., fine-to-coarse) for multi-scale representation with single-scale patches. In this work, with a different perspective from existing Transformers, we explore multi-scale patch embedding and multi-path structure, constructing the Multi-Path Vision Transformer (MPViT). MPViT embeds features of the same size~(i.e., sequence length) with patches of different scales simultaneously by using overlapping convolutional patch embedding. Tokens of different scales are then independently fed into the Transformer encoders via multiple paths and the resulting features are aggregated, enabling both fine and coarse feature representations at the same feature level. Thanks to the diverse, multi-scale feature representations, our MPViTs scaling from tiny~(5M) to base~(73M) consistently achieve superior performance over state-of-the-art Vision Transformers on ImageNet classification, object detection, instance segmentation, and semantic segmentation. These extensive results demonstrate that MPViT can serve as a versatile backbone network for various vision tasks. Code will be made publicly available at \url{https://git.io/MPViT}.
Video classification researches that have recently attracted attention are the fields of temporal modeling and 3D efficient architecture. However, the temporal modeling methods are not efficient or the 3D efficient architecture is less interested in temporal modeling. For bridging the gap between them, we propose an efficient temporal modeling 3D architecture, called VoV3D, that consists of a temporal one-shot aggregation (T-OSA) module and depthwise factorized component, D(2+1)D. The T-OSA is devised to build a feature hierarchy by aggregating temporal features with different temporal receptive fields. Stacking this T-OSA enables the network itself to model short-range as well as long-range temporal relationships across frames without any external modules. Inspired by kernel factorization and channel factorization, we also design a depthwise spatiotemporal factorization module, named, D(2+1)D that decomposes a 3D depthwise convolution into two spatial and temporal depthwise convolutions for making our network more lightweight and efficient. By using the proposed temporal modeling method (T-OSA), and the efficient factorized component (D(2+1)D), we construct two types of VoV3D networks, VoV3D-M and VoV3D-L. Thanks to its efficiency and effectiveness of temporal modeling, VoV3D-L has 6x fewer model parameters and 16x less computation, surpassing a state-of-the-art temporal modeling method on both Something-Something and Kinetics-400. Furthermore, VoV3D shows better temporal modeling ability than a state-of-the-art efficient 3D architecture, X3D having comparable model capacity. We hope that VoV3D can serve as a baseline for efficient video classification.
Adversarial training deep neural networks often experience serious overfitting problem. Recently, it is explained that the overfitting happens because the sample complexity of training data is insufficient to generalize robustness. In traditional machine learning, one way to relieve overfitting from the lack of data is to use ensemble methods. However, adversarial training multiple networks is extremely expensive. Moreover, we found that there is a dilemma on choosing target model to generate adversarial examples. Optimizing attack to the members of ensemble will be suboptimal attack to the ensemble and incurs covariate shift, while attack to ensemble will weaken the members and lose the benefit from ensembling. In this paper, we propose adversarial training with Stochastic weight average (SWA); while performing adversarial training, we aggregate the temporal weight states in the trajectory of training. By adopting SWA, the benefit of ensemble can be gained without tremendous computational increment and without facing the dilemma. Moreover, we further improved SWA to be adequate to adversarial training. The empirical results on CIFAR-10, CIFAR-100 and SVHN show that our method can improve the robustness of models.
Since many safety-critical systems such as surgical robots and autonomous driving cars are in unstable environments with sensor noise or incomplete data, it is desirable for object detectors to take the confidence of the localization prediction into account. Recent attempts to estimate localization uncertainty for object detection focus only anchor-based method that captures the uncertainty of different characteristics such as location (center point) and scale (width, height). Also, anchor-based methods need to adjust sensitive anchor-box settings. Therefore, we propose a new object detector called Gaussian-FCOS that estimates the localization uncertainty based on an anchor-free detector that captures the uncertainty of similar property with four directions of box offsets (left, right, top, bottom) and avoids the anchor tuning. For this purpose, we design a new loss function, uncertainty loss, to measure how uncertain the estimated object location is by modeling the uncertainty as a Gaussian distribution. Then, the detection score is calibrated through the estimated uncertainty. Experiments on challenging COCO datasets demonstrate that the proposed new loss function not only enables the network to estimate the uncertainty but produces a synergy effect with regression loss. In addition, our Gaussian-FCOS reduces false positives with the estimated localization uncertainty and finds more missing-objects, boosting both Average Precision (AP) and Recall (AR). We hope Gaussian-FCOS serve as a baseline for the reliability-required task.
We propose a simple yet efficient anchor-free instance segmentation, called CenterMask, that adds a novel spatial attention-guided mask (SAG-Mask) branch to anchor-free one stage object detector (FCOS) in the same vein with Mask R-CNN. Plugged into the FCOS object detector, the SAG-Mask branch predicts a segmentation mask on each box with the spatial attention map that helps to focus on informative pixels and suppress noise. We also present an improved VoVNetV2 with two effective strategies: adds (1) residual connection for alleviating the saturation problem of larger VoVNet and (2) effective Squeeze-Excitation (eSE) deals with the information loss problem of original SE. With SAG-Mask and VoVNetV2, we deign CenterMask and CenterMask-Lite that are targeted to large and small models, respectively. CenterMask outperforms all previous state-of-the-art models at a much faster speed. CenterMask-Lite also achieves 33.4\% mask AP / 38.0\% box AP, outperforming the state-of-the-art by 2.6 / 7.0 AP gain, respectively, at over 35fps on Titan Xp. We hope that CenterMask and VoVNetV2 can serve as a solid baseline of real-time instance segmentation and backbone network for various vision tasks, respectively. Code will be released.