Abhinav Shrivastava

Hao Chen, Matt Gwilliam, Ser-Nam Lim et al.

Implicit neural representations store videos as neural networks and have performed well for various vision tasks such as video compression and denoising. With frame index or positional index as input, implicit representations (NeRV, E-NeRV, \etc) reconstruct video from fixed and content-agnostic embeddings. Such embedding largely limits the regression capacity and internal generalization for video interpolation. In this paper, we propose a Hybrid Neural Representation for Videos (HNeRV), where a learnable encoder generates content-adaptive embeddings, which act as the decoder input. Besides the input embedding, we introduce HNeRV blocks, which ensure model parameters are evenly distributed across the entire network, such that higher layers (layers near the output) can have more capacity to store high-resolution content and video details. With content-adaptive embeddings and re-designed architecture, HNeRV outperforms implicit methods in video regression tasks for both reconstruction quality ($+4.7$ PSNR) and convergence speed ($16\times$ faster), and shows better internal generalization. As a simple and efficient video representation, HNeRV also shows decoding advantages for speed, flexibility, and deployment, compared to traditional codecs~(H.264, H.265) and learning-based compression methods. Finally, we explore the effectiveness of HNeRV on downstream tasks such as video compression and video inpainting. We provide project page at https://haochen-rye.github.io/HNeRV, and Code at https://github.com/haochen-rye/HNeRV

Bo He, Xitong Yang, Le Kang et al.

Weakly-supervised temporal action localization aims to recognize and localize action segments in untrimmed videos given only video-level action labels for training. Without the boundary information of action segments, existing methods mostly rely on multiple instance learning (MIL), where the predictions of unlabeled instances (i.e., video snippets) are supervised by classifying labeled bags (i.e., untrimmed videos). However, this formulation typically treats snippets in a video as independent instances, ignoring the underlying temporal structures within and across action segments. To address this problem, we propose \system, a novel WTAL framework that enables explicit, action-aware segment modeling beyond standard MIL-based methods. Our framework entails three segment-centric components: (i) dynamic segment sampling for compensating the contribution of short actions; (ii) intra- and inter-segment attention for modeling action dynamics and capturing temporal dependencies; (iii) pseudo instance-level supervision for improving action boundary prediction. Furthermore, a multi-step refinement strategy is proposed to progressively improve action proposals along the model training process. Extensive experiments on THUMOS-14 and ActivityNet-v1.3 demonstrate the effectiveness of our approach, establishing new state of the art on both datasets. The code and models are publicly available at~\url{https://github.com/boheumd/ASM-Loc}.

Kamal Gupta, Varun Jampani, Carlos Esteves et al. · deepmind

We present a method for joint alignment of sparse in-the-wild image collections of an object category. Most prior works assume either ground-truth keypoint annotations or a large dataset of images of a single object category. However, neither of the above assumptions hold true for the long-tail of the objects present in the world. We present a self-supervised technique that directly optimizes on a sparse collection of images of a particular object/object category to obtain consistent dense correspondences across the collection. We use pairwise nearest neighbors obtained from deep features of a pre-trained vision transformer (ViT) model as noisy and sparse keypoint matches and make them dense and accurate matches by optimizing a neural network that jointly maps the image collection into a learned canonical grid. Experiments on CUB and SPair-71k benchmarks demonstrate that our method can produce globally consistent and higher quality correspondences across the image collection when compared to existing self-supervised methods. Code and other material will be made available at \url{https://kampta.github.io/asic}.

Nirat Saini, Khoi Pham, Abhinav Shrivastava

We study the problem of compositional zero-shot learning for object-attribute recognition. Prior works use visual features extracted with a backbone network, pre-trained for object classification and thus do not capture the subtly distinct features associated with attributes. To overcome this challenge, these studies employ supervision from the linguistic space, and use pre-trained word embeddings to better separate and compose attribute-object pairs for recognition. Analogous to linguistic embedding space, which already has unique and agnostic embeddings for object and attribute, we shift the focus back to the visual space and propose a novel architecture that can disentangle attribute and object features in the visual space. We use visual decomposed features to hallucinate embeddings that are representative for the seen and novel compositions to better regularize the learning of our model. Extensive experiments show that our method outperforms existing work with significant margin on three datasets: MIT-States, UT-Zappos, and a new benchmark created based on VAW. The code, models, and dataset splits are publicly available at https://github.com/nirat1606/OADis.

Matthew Walmer, Saksham Suri, Kamal Gupta et al.

Vision Transformers (ViTs) have gained significant popularity in recent years and have proliferated into many applications. However, their behavior under different learning paradigms is not well explored. We compare ViTs trained through different methods of supervision, and show that they learn a diverse range of behaviors in terms of their attention, representations, and downstream performance. We also discover ViT behaviors that are consistent across supervision, including the emergence of Offset Local Attention Heads. These are self-attention heads that attend to a token adjacent to the current token with a fixed directional offset, a phenomenon that to the best of our knowledge has not been highlighted in any prior work. Our analysis shows that ViTs are highly flexible and learn to process local and global information in different orders depending on their training method. We find that contrastive self-supervised methods learn features that are competitive with explicitly supervised features, and they can even be superior for part-level tasks. We also find that the representations of reconstruction-based models show non-trivial similarity to contrastive self-supervised models. Project website (https://www.cs.umd.edu/~sakshams/vit_analysis) and code (https://www.github.com/mwalmer-umd/vit_analysis) are publicly available.

Sharath Girish, Kamal Gupta, Saurabh Singh et al.

We introduce LilNetX, an end-to-end trainable technique for neural networks that enables learning models with specified accuracy-rate-computation trade-off. Prior works approach these problems one at a time and often require post-processing or multistage training which become less practical and do not scale very well for large datasets or architectures. Our method constructs a joint training objective that penalizes the self-information of network parameters in a reparameterized latent space to encourage small model size while also introducing priors to increase structured sparsity in the parameter space to reduce computation. We achieve up to 50% smaller model size and 98% model sparsity on ResNet-20 while retaining the same accuracy on the CIFAR-10 dataset as well as 35% smaller model size and 42% structured sparsity on ResNet-50 trained on ImageNet, when compared to existing state-of-the-art model compression methods. Code is available at https://github.com/Sharath-girish/LilNetX.

Shishira R Maiya, Sharath Girish, Max Ehrlich et al. · nvidia

Implicit Neural Representations (INR) have recently shown to be powerful tool for high-quality video compression. However, existing works are limiting as they do not explicitly exploit the temporal redundancy in videos, leading to a long encoding time. Additionally, these methods have fixed architectures which do not scale to longer videos or higher resolutions. To address these issues, we propose NIRVANA, which treats videos as groups of frames and fits separate networks to each group performing patch-wise prediction. This design shares computation within each group, in the spatial and temporal dimensions, resulting in reduced encoding time of the video. The video representation is modeled autoregressively, with networks fit on a current group initialized using weights from the previous group's model. To further enhance efficiency, we perform quantization of the network parameters during training, requiring no post-hoc pruning or quantization. When compared with previous works on the benchmark UVG dataset, NIRVANA improves encoding quality from 37.36 to 37.70 (in terms of PSNR) and the encoding speed by 12X, while maintaining the same compression rate. In contrast to prior video INR works which struggle with larger resolution and longer videos, we show that our algorithm is highly flexible and scales naturally due to its patch-wise and autoregressive designs. Moreover, our method achieves variable bitrate compression by adapting to videos with varying inter-frame motion. NIRVANA achieves 6X decoding speed and scales well with more GPUs, making it practical for various deployment scenarios.

Junke Wang, Zuxuan Wu, Jingjing Chen et al.

Recent advances in image editing techniques have posed serious challenges to the trustworthiness of multimedia data, which drives the research of image tampering detection. In this paper, we propose ObjectFormer to detect and localize image manipulations. To capture subtle manipulation traces that are no longer visible in the RGB domain, we extract high-frequency features of the images and combine them with RGB features as multimodal patch embeddings. Additionally, we use a set of learnable object prototypes as mid-level representations to model the object-level consistencies among different regions, which are further used to refine patch embeddings to capture the patch-level consistencies. We conduct extensive experiments on various datasets and the results verify the effectiveness of the proposed method, outperforming state-of-the-art tampering detection and localization methods.

Sai Saketh Rambhatla, Ishan Misra, Rama Chellappa et al.

We tackle the challenging task of unsupervised object localization in this work. Recently, transformers trained with self-supervised learning have been shown to exhibit object localization properties without being trained for this task. In this work, we present Multiple Object localization with Self-supervised Transformers (MOST) that uses features of transformers trained using self-supervised learning to localize multiple objects in real world images. MOST analyzes the similarity maps of the features using box counting; a fractal analysis tool to identify tokens lying on foreground patches. The identified tokens are then clustered together, and tokens of each cluster are used to generate bounding boxes on foreground regions. Unlike recent state-of-the-art object localization methods, MOST can localize multiple objects per image and outperforms SOTA algorithms on several object localization and discovery benchmarks on PASCAL-VOC 07, 12 and COCO20k datasets. Additionally, we show that MOST can be used for self-supervised pre-training of object detectors, and yields consistent improvements on fully, semi-supervised object detection and unsupervised region proposal generation.

Bo He, Jun Wang, Jielin Qiu et al.

The goal of multimodal summarization is to extract the most important information from different modalities to form output summaries. Unlike the unimodal summarization, the multimodal summarization task explicitly leverages cross-modal information to help generate more reliable and high-quality summaries. However, existing methods fail to leverage the temporal correspondence between different modalities and ignore the intrinsic correlation between different samples. To address this issue, we introduce Align and Attend Multimodal Summarization (A2Summ), a unified multimodal transformer-based model which can effectively align and attend the multimodal input. In addition, we propose two novel contrastive losses to model both inter-sample and intra-sample correlations. Extensive experiments on two standard video summarization datasets (TVSum and SumMe) and two multimodal summarization datasets (Daily Mail and CNN) demonstrate the superiority of A2Summ, achieving state-of-the-art performances on all datasets. Moreover, we collected a large-scale multimodal summarization dataset BLiSS, which contains livestream videos and transcribed texts with annotated summaries. Our code and dataset are publicly available at ~\url{https://boheumd.github.io/A2Summ/}.

Bo He, Xitong Yang, Hanyu Wang et al.

Implicit neural representations (INR) have gained increasing attention in representing 3D scenes and images, and have been recently applied to encode videos (e.g., NeRV, E-NeRV). While achieving promising results, existing INR-based methods are limited to encoding a handful of short videos (e.g., seven 5-second videos in the UVG dataset) with redundant visual content, leading to a model design that fits individual video frames independently and is not efficiently scalable to a large number of diverse videos. This paper focuses on developing neural representations for a more practical setup -- encoding long and/or a large number of videos with diverse visual content. We first show that instead of dividing videos into small subsets and encoding them with separate models, encoding long and diverse videos jointly with a unified model achieves better compression results. Based on this observation, we propose D-NeRV, a novel neural representation framework designed to encode diverse videos by (i) decoupling clip-specific visual content from motion information, (ii) introducing temporal reasoning into the implicit neural network, and (iii) employing the task-oriented flow as intermediate output to reduce spatial redundancies. Our new model largely surpasses NeRV and traditional video compression techniques on UCF101 and UVG datasets on the video compression task. Moreover, when used as an efficient data-loader, D-NeRV achieves 3%-10% higher accuracy than NeRV on action recognition tasks on the UCF101 dataset under the same compression ratios.

Soumik Mukhopadhyay, Matthew Gwilliam, Vatsal Agarwal et al.

While many unsupervised learning models focus on one family of tasks, either generative or discriminative, we explore the possibility of a unified representation learner: a model which uses a single pre-training stage to address both families of tasks simultaneously. We identify diffusion models as a prime candidate. Diffusion models have risen to prominence as a state-of-the-art method for image generation, denoising, inpainting, super-resolution, manipulation, etc. Such models involve training a U-Net to iteratively predict and remove noise, and the resulting model can synthesize high fidelity, diverse, novel images. The U-Net architecture, as a convolution-based architecture, generates a diverse set of feature representations in the form of intermediate feature maps. We present our findings that these embeddings are useful beyond the noise prediction task, as they contain discriminative information and can also be leveraged for classification. We explore optimal methods for extracting and using these embeddings for classification tasks, demonstrating promising results on the ImageNet classification task. We find that with careful feature selection and pooling, diffusion models outperform comparable generative-discriminative methods such as BigBiGAN for classification tasks. We investigate diffusion models in the transfer learning regime, examining their performance on several fine-grained visual classification datasets. We compare these embeddings to those generated by competing architectures and pre-trainings for classification tasks.

Yifei Zhou, Zilu Li, Abhinav Shrivastava et al.

Modern retrieval system often requires recomputing the representation of every piece of data in the gallery when updating to a better representation model. This process is known as backfilling and can be especially costly in the real world where the gallery often contains billions of samples. Recently, researchers have proposed the idea of Backward Compatible Training (BCT) where the new representation model can be trained with an auxiliary loss to make it backward compatible with the old representation. In this way, the new representation can be directly compared with the old representation, in principle avoiding the need for any backfilling. However, followup work shows that there is an inherent tradeoff where a backward compatible representation model cannot simultaneously maintain the performance of the new model itself. This paper reports our ``not-so-surprising'' finding that adding extra dimensions to the representation can help here. However, we also found that naively increasing the dimension of the representation did not work. To deal with this, we propose Backward-compatible Training with a novel Basis Transformation ($BT^2$). A basis transformation (BT) is basically a learnable set of parameters that applies an orthonormal transformation. Such a transformation possesses an important property whereby the original information contained in its input is retained in its output. We show in this paper how a BT can be utilized to add only the necessary amount of additional dimensions. We empirically verify the advantage of $BT^2$ over other state-of-the-art methods in a wide range of settings. We then further extend $BT^2$ to other challenging yet more practical settings, including significant change in model architecture (CNN to Transformers), modality change, and even a series of updates in the model architecture mimicking the evolution of deep learning models.

Soumik Mukhopadhyay, Saksham Suri, Ravi Teja Gadde et al.

The task of lip synchronization (lip-sync) seeks to match the lips of human faces with different audio. It has various applications in the film industry as well as for creating virtual avatars and for video conferencing. This is a challenging problem as one needs to simultaneously introduce detailed, realistic lip movements while preserving the identity, pose, emotions, and image quality. Many of the previous methods trying to solve this problem suffer from image quality degradation due to a lack of complete contextual information. In this paper, we present Diff2Lip, an audio-conditioned diffusion-based model which is able to do lip synchronization in-the-wild while preserving these qualities. We train our model on Voxceleb2, a video dataset containing in-the-wild talking face videos. Extensive studies show that our method outperforms popular methods like Wav2Lip and PC-AVS in Fréchet inception distance (FID) metric and Mean Opinion Scores (MOS) of the users. We show results on both reconstruction (same audio-video inputs) as well as cross (different audio-video inputs) settings on Voxceleb2 and LRW datasets. Video results and code can be accessed from our project page ( https://soumik-kanad.github.io/diff2lip ).

Shuaiyi Huang, Luyu Yang, Bo He et al.

Finding dense semantic correspondence is a fundamental problem in computer vision, which remains challenging in complex scenes due to background clutter, extreme intra-class variation, and a severe lack of ground truth. In this paper, we aim to address the challenge of label sparsity in semantic correspondence by enriching supervision signals from sparse keypoint annotations. To this end, we first propose a teacher-student learning paradigm for generating dense pseudo-labels and then develop two novel strategies for denoising pseudo-labels. In particular, we use spatial priors around the sparse annotations to suppress the noisy pseudo-labels. In addition, we introduce a loss-driven dynamic label selection strategy for label denoising. We instantiate our paradigm with two variants of learning strategies: a single offline teacher setting, and mutual online teachers setting. Our approach achieves notable improvements on three challenging benchmarks for semantic correspondence and establishes the new state-of-the-art. Project page: https://shuaiyihuang.github.io/publications/SCorrSAN.

Fuxiao Liu, Yaser Yacoob, Abhinav Shrivastava

We introduce a new benchmark, COVID-VTS, for fact-checking multi-modal information involving short-duration videos with COVID19- focused information from both the real world and machine generation. We propose, TwtrDetective, an effective model incorporating cross-media consistency checking to detect token-level malicious tampering in different modalities, and generate explanations. Due to the scarcity of training data, we also develop an efficient and scalable approach to automatically generate misleading video posts by event manipulation or adversarial matching. We investigate several state-of-the-art models and demonstrate the superiority of TwtrDetective.

Hao Chen, Matt Gwilliam, Bo He et al.

Compression and reconstruction of visual data have been widely studied in the computer vision community, even before the popularization of deep learning. More recently, some have used deep learning to improve or refine existing pipelines, while others have proposed end-to-end approaches, including autoencoders and implicit neural representations, such as SIREN and NeRV. In this work, we propose Neural Visual Representation with Content-adaptive Embedding (CNeRV), which combines the generalizability of autoencoders with the simplicity and compactness of implicit representation. We introduce a novel content-adaptive embedding that is unified, concise, and internally (within-video) generalizable, that compliments a powerful decoder with a single-layer encoder. We match the performance of NeRV, a state-of-the-art implicit neural representation, on the reconstruction task for frames seen during training while far surpassing for frames that are skipped during training (unseen images). To achieve similar reconstruction quality on unseen images, NeRV needs 120x more time to overfit per-frame due to its lack of internal generalization. With the same latent code length and similar model size, CNeRV outperforms autoencoders on reconstruction of both seen and unseen images. We also show promising results for visual data compression. More details can be found in the project pagehttps://haochen-rye.github.io/CNeRV/

Sharath Girish, Abhinav Shrivastava, Kamal Gupta

Implicit Neural Representations (INR) or neural fields have emerged as a popular framework to encode multimedia signals such as images and radiance fields while retaining high-quality. Recently, learnable feature grids proposed by Instant-NGP have allowed significant speed-up in the training as well as the sampling of INRs by replacing a large neural network with a multi-resolution look-up table of feature vectors and a much smaller neural network. However, these feature grids come at the expense of large memory consumption which can be a bottleneck for storage and streaming applications. In this work, we propose SHACIRA, a simple yet effective task-agnostic framework for compressing such feature grids with no additional post-hoc pruning/quantization stages. We reparameterize feature grids with quantized latent weights and apply entropy regularization in the latent space to achieve high levels of compression across various domains. Quantitative and qualitative results on diverse datasets consisting of images, videos, and radiance fields, show that our approach outperforms existing INR approaches without the need for any large datasets or domain-specific heuristics. Our project page is available at http://shacira.github.io .

Matthew Gwilliam, Abhinav Shrivastava

By leveraging contrastive learning, clustering, and other pretext tasks, unsupervised methods for learning image representations have reached impressive results on standard benchmarks. The result has been a crowded field - many methods with substantially different implementations yield results that seem nearly identical on popular benchmarks, such as linear evaluation on ImageNet. However, a single result does not tell the whole story. In this paper, we compare methods using performance-based benchmarks such as linear evaluation, nearest neighbor classification, and clustering for several different datasets, demonstrating the lack of a clear front-runner within the current state-of-the-art. In contrast to prior work that performs only supervised vs. unsupervised comparison, we compare several different unsupervised methods against each other. To enrich this comparison, we analyze embeddings with measurements such as uniformity, tolerance, and centered kernel alignment (CKA), and propose two new metrics of our own: nearest neighbor graph similarity and linear prediction overlap. We reveal through our analysis that in isolation, single popular methods should not be treated as though they represent the field as a whole, and that future work ought to consider how to leverage the complimentary nature of these methods. We also leverage CKA to provide a framework to robustly quantify augmentation invariance, and provide a reminder that certain types of invariance will be undesirable for downstream tasks.

Vinoj Jayasundara, Amit Agrawal, Nicolas Heron et al.

We present FlexNeRF, a method for photorealistic freeviewpoint rendering of humans in motion from monocular videos. Our approach works well with sparse views, which is a challenging scenario when the subject is exhibiting fast/complex motions. We propose a novel approach which jointly optimizes a canonical time and pose configuration, with a pose-dependent motion field and pose-independent temporal deformations complementing each other. Thanks to our novel temporal and cyclic consistency constraints along with additional losses on intermediate representation such as segmentation, our approach provides high quality outputs as the observed views become sparser. We empirically demonstrate that our method significantly outperforms the state-of-the-art on public benchmark datasets as well as a self-captured fashion dataset. The project page is available at: https://flex-nerf.github.io/

Nirat Saini, Hanyu Wang, Archana Swaminathan et al.

Recognizing and generating object-state compositions has been a challenging task, especially when generalizing to unseen compositions. In this paper, we study the task of cutting objects in different styles and the resulting object state changes. We propose a new benchmark suite Chop & Learn, to accommodate the needs of learning objects and different cut styles using multiple viewpoints. We also propose a new task of Compositional Image Generation, which can transfer learned cut styles to different objects, by generating novel object-state images. Moreover, we also use the videos for Compositional Action Recognition, and show valuable uses of this dataset for multiple video tasks. Project website: https://chopnlearn.github.io.

Pulkit Kumar, Namitha Padmanabhan, Luke Luo et al.

We propose a simple yet effective approach for few-shot action recognition, emphasizing the disentanglement of motion and appearance representations. By harnessing recent progress in tracking, specifically point trajectories and self-supervised representation learning, we build trajectory-aligned tokens (TATs) that capture motion and appearance information. This approach significantly reduces the data requirements while retaining essential information. To process these representations, we use a Masked Space-time Transformer that effectively learns to aggregate information to facilitate few-shot action recognition. We demonstrate state-of-the-art results on few-shot action recognition across multiple datasets. Our project page is available at https://www.cs.umd.edu/~pulkit/tats

Nirat Saini, Navaneeth Bodla, Ashish Shrivastava et al.

We introduce InVi, an approach for inserting or replacing objects within videos (referred to as inpainting) using off-the-shelf, text-to-image latent diffusion models. InVi targets controlled manipulation of objects and blending them seamlessly into a background video unlike existing video editing methods that focus on comprehensive re-styling or entire scene alterations. To achieve this goal, we tackle two key challenges. Firstly, for high quality control and blending, we employ a two-step process involving inpainting and matching. This process begins with inserting the object into a single frame using a ControlNet-based inpainting diffusion model, and then generating subsequent frames conditioned on features from an inpainted frame as an anchor to minimize the domain gap between the background and the object. Secondly, to ensure temporal coherence, we replace the diffusion model's self-attention layers with extended-attention layers. The anchor frame features serve as the keys and values for these layers, enhancing consistency across frames. Our approach removes the need for video-specific fine-tuning, presenting an efficient and adaptable solution. Experimental results demonstrate that InVi achieves realistic object insertion with consistent blending and coherence across frames, outperforming existing methods.

Shishira R Maiya, Anubhav Gupta, Matthew Gwilliam et al.

Implicit Neural Networks (INRs) have emerged as powerful representations to encode all forms of data, including images, videos, audios, and scenes. With video, many INRs for video have been proposed for the compression task, and recent methods feature significant improvements with respect to encoding time, storage, and reconstruction quality. However, these encoded representations lack semantic meaning, so they cannot be used for any downstream tasks that require such properties, such as retrieval. This can act as a barrier for adoption of video INRs over traditional codecs as they do not offer any significant edge apart from compression. To alleviate this, we propose a flexible framework that decouples the spatial and temporal aspects of the video INR. We accomplish this with a dictionary of per-frame latents that are learned jointly with a set of video specific hypernetworks, such that given a latent, these hypernetworks can predict the INR weights to reconstruct the given frame. This framework not only retains the compression efficiency, but the learned latents can be aligned with features from large vision models, which grants them discriminative properties. We align these latents with CLIP and show good performance for both compression and video retrieval tasks. By aligning with VideoLlama, we are able to perform open-ended chat with our learned latents as the visual inputs. Additionally, the learned latents serve as a proxy for the underlying weights, allowing us perform tasks like video interpolation. These semantic properties and applications, existing simultaneously with ability to perform compression, interpolation, and superresolution properties, are a first in this field of work.

Mara Levy, Abhinav Shrivastava

Learning to represent three dimensional (3D) human pose given a two dimensional (2D) image of a person, is a challenging problem. In order to make the problem less ambiguous it has become common practice to estimate 3D pose in the camera coordinate space. However, this makes the task of comparing two 3D poses difficult. In this paper, we address this challenge by separating the problem of estimating 3D pose from 2D images into two steps. We use a variational autoencoder (VAE) to find an embedding that represents 3D poses in canonical coordinate space. We refer to this embedding as variational view-invariant pose embedding V-VIPE. Using V-VIPE we can encode 2D and 3D poses and use the embedding for downstream tasks, like retrieval and classification. We can estimate 3D poses from these embeddings using the decoder as well as generate unseen 3D poses. The variability of our encoding allows it to generalize well to unseen camera views when mapping from 2D space. To the best of our knowledge, V-VIPE is the only representation to offer this diversity of applications. Code and more information can be found at https://v-vipe.github.io/.

Hanyu Wang, Kamal Gupta, Larry Davis et al.

We present Neural Space-filling Curves (SFCs), a data-driven approach to infer a context-based scan order for a set of images. Linear ordering of pixels forms the basis for many applications such as video scrambling, compression, and auto-regressive models that are used in generative modeling for images. Existing algorithms resort to a fixed scanning algorithm such as Raster scan or Hilbert scan. Instead, our work learns a spatially coherent linear ordering of pixels from the dataset of images using a graph-based neural network. The resulting Neural SFC is optimized for an objective suitable for the downstream task when the image is traversed along with the scan line order. We show the advantage of using Neural SFCs in downstream applications such as image compression. Code and additional results will be made available at https://hywang66.github.io/publication/neuralsfc.

Mara Levy, Nirat Saini, Abhinav Shrivastava

We propose WayEx, a new method for learning complex goal-conditioned robotics tasks from a single demonstration. Our approach distinguishes itself from existing imitation learning methods by demanding fewer expert examples and eliminating the need for information about the actions taken during the demonstration. This is accomplished by introducing a new reward function and employing a knowledge expansion technique. We demonstrate the effectiveness of WayEx, our waypoint exploration strategy, across six diverse tasks, showcasing its applicability in various environments. Notably, our method significantly reduces training time by 50% as compared to traditional reinforcement learning methods. WayEx obtains a higher reward than existing imitation learning methods given only a single demonstration. Furthermore, we demonstrate its success in tackling complex environments where standard approaches fall short. More information is available at: https://waypoint-ex.github.io.

Sharath Girish, Debadeepta Dey, Neel Joshi et al.

The current literature on self-supervised learning (SSL) focuses on developing learning objectives to train neural networks more effectively on unlabeled data. The typical development process involves taking well-established architectures, e.g., ResNet demonstrated on ImageNet, and using them to evaluate newly developed objectives on downstream scenarios. While convenient, this does not take into account the role of architectures which has been shown to be crucial in the supervised learning literature. In this work, we establish extensive empirical evidence showing that a network architecture plays a significant role in SSL. We conduct a large-scale study with over 100 variants of ResNet and MobileNet architectures and evaluate them across 11 downstream scenarios in the SSL setting. We show that there is no one network that performs consistently well across the scenarios. Based on this, we propose to learn not only network weights but also architecture topologies in the SSL regime. We show that "self-supervised architectures" outperform popular handcrafted architectures (ResNet18 and MobileNetV2) while performing competitively with the larger and computationally heavy ResNet50 on major image classification benchmarks (ImageNet-1K, iNat2021, and more). Our results suggest that it is time to consider moving beyond handcrafted architectures in SSL and start thinking about incorporating architecture search into self-supervised learning objectives.

Matthew Gwilliam, Michael Cogswell, Meng Ye et al.

Existing long video retrieval systems are trained and tested in the paragraph-to-video retrieval regime, where every long video is described by a single long paragraph. This neglects the richness and variety of possible valid descriptions of a video, which could range anywhere from moment-by-moment detail to a single phrase summary. To provide a more thorough evaluation of the capabilities of long video retrieval systems, we propose a pipeline that leverages state-of-the-art large language models to carefully generate a diverse set of synthetic captions for long videos. We validate this pipeline's fidelity via rigorous human inspection. We use synthetic captions from this pipeline to perform a benchmark of a representative set of video language models using long video datasets, and show that the models struggle on shorter captions. We show that finetuning on this data can both mitigate these issues (+2.8% R@1 over SOTA on ActivityNet with diverse captions), and even improve performance on standard paragraph-to-video retrieval (+1.0% R@1 on ActivityNet). We also use synthetic data from our pipeline as query expansion in the zero-shot setting (+3.4% R@1 on ActivityNet). We derive insights by analyzing failure cases for retrieval with short captions. For data access and other details, please refer to our project website at https://mgwillia.github.io/10k-words.

Bo He, Hengduo Li, Young Kyun Jang et al.

With the success of large language models (LLMs), integrating the vision model into LLMs to build vision-language foundation models has gained much more interest recently. However, existing LLM-based large multimodal models (e.g., Video-LLaMA, VideoChat) can only take in a limited number of frames for short video understanding. In this study, we mainly focus on designing an efficient and effective model for long-term video understanding. Instead of trying to process more frames simultaneously like most existing work, we propose to process videos in an online manner and store past video information in a memory bank. This allows our model to reference historical video content for long-term analysis without exceeding LLMs' context length constraints or GPU memory limits. Our memory bank can be seamlessly integrated into current multimodal LLMs in an off-the-shelf manner. We conduct extensive experiments on various video understanding tasks, such as long-video understanding, video question answering, and video captioning, and our model can achieve state-of-the-art performances across multiple datasets. Code available at https://boheumd.github.io/MA-LMM/.

Sharath Girish, Kamal Gupta, Abhinav Shrivastava

Recently, 3D Gaussian splatting (3D-GS) has gained popularity in novel-view scene synthesis. It addresses the challenges of lengthy training times and slow rendering speeds associated with Neural Radiance Fields (NeRFs). Through rapid, differentiable rasterization of 3D Gaussians, 3D-GS achieves real-time rendering and accelerated training. They, however, demand substantial memory resources for both training and storage, as they require millions of Gaussians in their point cloud representation for each scene. We present a technique utilizing quantized embeddings to significantly reduce per-point memory storage requirements and a coarse-to-fine training strategy for a faster and more stable optimization of the Gaussian point clouds. Our approach develops a pruning stage which results in scene representations with fewer Gaussians, leading to faster training times and rendering speeds for real-time rendering of high resolution scenes. We reduce storage memory by more than an order of magnitude all while preserving the reconstruction quality. We validate the effectiveness of our approach on a variety of datasets and scenes preserving the visual quality while consuming 10-20x lesser memory and faster training/inference speed. Project page and code is available https://efficientgaussian.github.io

Matthew Walmer, Rose Kanjirathinkal, Kai Sheng Tai et al.

The area of temporally fine-grained video representation learning focuses on generating frame-by-frame representations for temporally dense tasks, such as fine-grained action phase classification and frame retrieval. In this work, we advance the state-of-the-art for self-supervised models in this area by re-examining the design of transformer architectures for video representation learning. A key aspect of our approach is the improved sharing of scene information in the temporal pipeline by representing multiple salient entities per frame. Prior works use late-fusion architectures that reduce frames to a single-dimensional vector before modeling any cross-frame dynamics. In contrast, our Multi-entity Video Transformer (MV-Former) processes the frames as groups of entities represented as tokens linked across time. To achieve this, we propose a Learnable Spatial Token Pooling strategy to identify and extract features for multiple salient regions per frame. Through our experiments, we show that MV-Former outperforms previous self-supervised methods, and also surpasses some prior works that use additional supervision or training data. When combined with additional pre-training data from Kinetics-400, MV-Former achieves a further performance boost. Overall, our MV-Former achieves state-of-the-art results on multiple fine-grained video benchmarks and shows that parsing video scenes as collections of entities can enhance performance in video tasks.

Gowthami Somepalli, Anubhav Gupta, Kamal Gupta et al. · microsoft-research

Generative models are now widely used by graphic designers and artists. Prior works have shown that these models remember and often replicate content from their training data during generation. Hence as their proliferation increases, it has become important to perform a database search to determine whether the properties of the image are attributable to specific training data, every time before a generated image is used for professional purposes. Existing tools for this purpose focus on retrieving images of similar semantic content. Meanwhile, many artists are concerned with style replication in text-to-image models. We present a framework for understanding and extracting style descriptors from images. Our framework comprises a new dataset curated using the insight that style is a subjective property of an image that captures complex yet meaningful interactions of factors including but not limited to colors, textures, shapes, etc. We also propose a method to extract style descriptors that can be used to attribute style of a generated image to the images used in the training dataset of a text-to-image model. We showcase promising results in various style retrieval tasks. We also quantitatively and qualitatively analyze style attribution and matching in the Stable Diffusion model. Code and artifacts are available at https://github.com/learn2phoenix/CSD.

Hao Chen, Saining Xie, Ser-Nam Lim et al.

Despite the abundant availability and content richness for video data, its high-dimensionality poses challenges for video research. Recent advancements have explored the implicit representation for videos using neural networks, demonstrating strong performance in applications such as video compression and enhancement. However, the prolonged encoding time remains a persistent challenge for video Implicit Neural Representations (INRs). In this paper, we focus on improving the speed of video encoding and decoding within implicit representations. We introduce two key components: NeRV-Enc, a transformer-based hyper-network for fast encoding; and NeRV-Dec, a parallel decoder for efficient video loading. NeRV-Enc achieves an impressive speed-up of $\mathbf{10^4\times}$ by eliminating gradient-based optimization. Meanwhile, NeRV-Dec simplifies video decoding, outperforming conventional codecs with a loading speed $\mathbf{11\times}$ faster, and surpassing RAM loading with pre-decoded videos ($\mathbf{2.5\times}$ faster while being $\mathbf{65\times}$ smaller in size).

Anirud Aggarwal, Abhinav Shrivastava, Matthew Gwilliam

Diffusion-based image generation models excel at producing high-quality synthetic content, but suffer from slow and computationally expensive inference. Prior work has attempted to mitigate this by caching and reusing features within diffusion transformers across inference steps. These methods, however, often rely on rigid heuristics that result in limited acceleration or poor generalization across architectures. We propose Evolutionary Caching to Accelerate Diffusion models (ECAD), a genetic algorithm that learns efficient, per-model, caching schedules forming a Pareto frontier, using only a small set of calibration prompts. ECAD requires no modifications to network parameters or reference images. It offers significant inference speedups, enables fine-grained control over the quality-latency trade-off, and adapts seamlessly to different diffusion models. Notably, ECAD's learned schedules can generalize effectively to resolutions and model variants not seen during calibration. We evaluate ECAD on PixArt-alpha, PixArt-Sigma, and FLUX-1$.$dev using multiple metrics (FID, CLIP, Image Reward) across diverse benchmarks (COCO, MJHQ-30k, PartiPrompts), demonstrating consistent improvements over previous approaches. On PixArt-alpha, ECAD identifies a schedule that outperforms the previous state-of-the-art method by 4.47 COCO FID while increasing inference speedup from 2.35x to 2.58x. Our results establish ECAD as a scalable and generalizable approach for accelerating diffusion inference. Our project website is available at https://aniaggarwal.github.io/ecad and our code is available at https://github.com/aniaggarwal/ecad.

Xiyang Wu, Tianrui Guan, Dianqi Li et al.

Large vision-language models (LVLMs) are prone to hallucinations, where certain contextual cues in an image can trigger the language module to produce overconfident and incorrect reasoning about abnormal or hypothetical objects. While some benchmarks have been developed to investigate LVLM hallucinations, they often rely on hand-crafted corner cases whose failure patterns may not generalize well. Additionally, fine-tuning on these examples could undermine their validity. To address this, we aim to scale up the number of cases through an automated approach, reducing human bias in crafting such corner cases. This motivates the development of AutoHallusion, the first automated benchmark generation approach that employs several key strategies to create a diverse range of hallucination examples. Our generated visual-question pairs pose significant challenges to LVLMs, requiring them to overcome contextual biases and distractions to arrive at correct answers. AutoHallusion enables us to create new benchmarks at the minimum cost and thus overcomes the fragility of hand-crafted benchmarks. It also reveals common failure patterns and reasons, providing key insights to detect, avoid, or control hallucinations. Comprehensive evaluations of top-tier LVLMs, e.g., GPT-4V(ision), Gemini Pro Vision, Claude 3, and LLaVA-1.5, show a 97.7% and 98.7% success rate of hallucination induction on synthetic and real-world datasets of AutoHallusion, paving the way for a long battle against hallucinations. The codebase and data can be accessed at https://github.com/wuxiyang1996/AutoHallusion.

Kamal Gupta, Gowthami Somepalli, Anubhav Gupta et al.

We study a referential game (a type of signaling game) where two agents communicate with each other via a discrete bottleneck to achieve a common goal. In our referential game, the goal of the speaker is to compose a message or a symbolic representation of "important" image patches, while the task for the listener is to match the speaker's message to a different view of the same image. We show that it is indeed possible for the two agents to develop a communication protocol without explicit or implicit supervision. We further investigate the developed protocol and show the applications in speeding up recent Vision Transformers by using only important patches, and as pre-training for downstream recognition tasks (e.g., classification). Code available at https://github.com/kampta/PatchGame.

Hao Chen, Bo He, Hanyu Wang et al.

We propose a novel neural representation for videos (NeRV) which encodes videos in neural networks. Unlike conventional representations that treat videos as frame sequences, we represent videos as neural networks taking frame index as input. Given a frame index, NeRV outputs the corresponding RGB image. Video encoding in NeRV is simply fitting a neural network to video frames and decoding process is a simple feedforward operation. As an image-wise implicit representation, NeRV output the whole image and shows great efficiency compared to pixel-wise implicit representation, improving the encoding speed by 25x to 70x, the decoding speed by 38x to 132x, while achieving better video quality. With such a representation, we can treat videos as neural networks, simplifying several video-related tasks. For example, conventional video compression methods are restricted by a long and complex pipeline, specifically designed for the task. In contrast, with NeRV, we can use any neural network compression method as a proxy for video compression, and achieve comparable performance to traditional frame-based video compression approaches (H.264, HEVC \etc). Besides compression, we demonstrate the generalization of NeRV for video denoising. The source code and pre-trained model can be found at https://github.com/haochen-rye/NeRV.git.

Ahmed Taha, Abhinav Shrivastava, Larry Davis

Deep learning relies on the availability of a large corpus of data (labeled or unlabeled). Thus, one challenging unsettled question is: how to train a deep network on a relatively small dataset? To tackle this question, we propose an evolution-inspired training approach to boost performance on relatively small datasets. The knowledge evolution (KE) approach splits a deep network into two hypotheses: the fit-hypothesis and the reset-hypothesis. We iteratively evolve the knowledge inside the fit-hypothesis by perturbing the reset-hypothesis for multiple generations. This approach not only boosts performance, but also learns a slim network with a smaller inference cost. KE integrates seamlessly with both vanilla and residual convolutional networks. KE reduces both overfitting and the burden for data collection. We evaluate KE on various network architectures and loss functions. We evaluate KE using relatively small datasets (e.g., CUB-200) and randomly initialized deep networks. KE achieves an absolute 21% improvement margin on a state-of-the-art baseline. This performance improvement is accompanied by a relative 73% reduction in inference cost. KE achieves state-of-the-art results on classification and metric learning benchmarks. Code available at http://bit.ly/3uLgwYb

Ahmed Taha, Alex Hanson, Abhinav Shrivastava et al.

A neural network regularizer (e.g., weight decay) boosts performance by explicitly penalizing the complexity of a network. In this paper, we penalize inferior network activations -- feature embeddings -- which in turn regularize the network's weights implicitly. We propose singular value maximization (SVMax) to learn a more uniform feature embedding. The SVMax regularizer supports both supervised and unsupervised learning. Our formulation mitigates model collapse and enables larger learning rates. We evaluate the SVMax regularizer using both retrieval and generative adversarial networks. We leverage a synthetic mixture of Gaussians dataset to evaluate SVMax in an unsupervised setting. For retrieval networks, SVMax achieves significant improvement margins across various ranking losses. Code available at https://bit.ly/3jNkgDt

Luyu Yang, Yan Wang, Mingfei Gao et al.

Semi-supervised domain adaptation (SSDA) aims to adapt models trained from a labeled source domain to a different but related target domain, from which unlabeled data and a small set of labeled data are provided. Current methods that treat source and target supervision without distinction overlook their inherent discrepancy, resulting in a source-dominated model that has not effectively used the target supervision. In this paper, we argue that the labeled target data needs to be distinguished for effective SSDA, and propose to explicitly decompose the SSDA task into two sub-tasks: a semi-supervised learning (SSL) task in the target domain and an unsupervised domain adaptation (UDA) task across domains. By doing so, the two sub-tasks can better leverage the corresponding supervision and thus yield very different classifiers. To integrate the strengths of the two classifiers, we apply the well-established co-training framework, in which the two classifiers exchange their high confident predictions to iteratively "teach each other" so that both classifiers can excel in the target domain. We call our approach Deep Co-training with Task decomposition (DeCoTa). DeCoTa requires no adversarial training and is easy to implement. Moreover, DeCoTa is well-founded on the theoretical condition of when co-training would succeed. As a result, DeCoTa achieves state-of-the-art results on several SSDA datasets, outperforming the prior art by a notable 4% margin on DomainNet. Code is available at https://github.com/LoyoYang/DeCoTa

Ahmed Taha, Xitong Yang, Abhinav Shrivastava et al.

Retrieval networks are essential for searching and indexing. Compared to classification networks, attention visualization for retrieval networks is hardly studied. We formulate attention visualization as a constrained optimization problem. We leverage the unit L2-Norm constraint as an attention filter (L2-CAF) to localize attention in both classification and retrieval networks. Unlike recent literature, our approach requires neither architectural changes nor fine-tuning. Thus, a pre-trained network's performance is never undermined L2-CAF is quantitatively evaluated using weakly supervised object localization. State-of-the-art results are achieved on classification networks. For retrieval networks, significant improvement margins are achieved over a Grad-CAM baseline. Qualitative evaluation demonstrates how the L2-CAF visualizes attention per frame for a recurrent retrieval network. Further ablation studies highlight the computational cost of our approach and compare L2-CAF with other feasible alternatives. Code available at https://bit.ly/3iDBLFv

Ahmed Taha, Yi-Ting Chen, Teruhisa Misu et al.

We employ triplet loss as a feature embedding regularizer to boost classification performance. Standard architectures, like ResNet and Inception, are extended to support both losses with minimal hyper-parameter tuning. This promotes generality while fine-tuning pretrained networks. Triplet loss is a powerful surrogate for recently proposed embedding regularizers. Yet, it is avoided due to large batch-size requirement and high computational cost. Through our experiments, we re-assess these assumptions. During inference, our network supports both classification and embedding tasks without any computational overhead. Quantitative evaluation highlights a steady improvement on five fine-grained recognition datasets. Further evaluation on an imbalanced video dataset achieves significant improvement. Triplet loss brings feature embedding characteristics like nearest neighbor to classification models. Code available at \url{http://bit.ly/2LNYEqL}.

Yixuan Ren, Yang Zhou, Jimei Yang et al.

Image customization has been extensively studied in text-to-image (T2I) diffusion models, leading to impressive outcomes and applications. With the emergence of text-to-video (T2V) diffusion models, its temporal counterpart, motion customization, has not yet been well investigated. To address the challenge of one-shot video motion customization, we propose Customize-A-Video that models the motion from a single reference video and adapts it to new subjects and scenes with both spatial and temporal varieties. It leverages low-rank adaptation (LoRA) on temporal attention layers to tailor the pre-trained T2V diffusion model for specific motion modeling. To disentangle the spatial and temporal information during training, we introduce a novel concept of appearance absorbers that detach the original appearance from the reference video prior to motion learning. The proposed modules are trained in a staged pipeline and inferred in a plug-and-play fashion, enabling easy extensions to various downstream tasks such as custom video generation and editing, video appearance customization and multiple motion combination. Our project page can be found at https://customize-a-video.github.io.

Hanyu Wang, Saksham Suri, Yixuan Ren et al.

We present LARP, a novel video tokenizer designed to overcome limitations in current video tokenization methods for autoregressive (AR) generative models. Unlike traditional patchwise tokenizers that directly encode local visual patches into discrete tokens, LARP introduces a holistic tokenization scheme that gathers information from the visual content using a set of learned holistic queries. This design allows LARP to capture more global and semantic representations, rather than being limited to local patch-level information. Furthermore, it offers flexibility by supporting an arbitrary number of discrete tokens, enabling adaptive and efficient tokenization based on the specific requirements of the task. To align the discrete token space with downstream AR generation tasks, LARP integrates a lightweight AR transformer as a training-time prior model that predicts the next token on its discrete latent space. By incorporating the prior model during training, LARP learns a latent space that is not only optimized for video reconstruction but is also structured in a way that is more conducive to autoregressive generation. Moreover, this process defines a sequential order for the discrete tokens, progressively pushing them toward an optimal configuration during training, ensuring smoother and more accurate AR generation at inference time. Comprehensive experiments demonstrate LARP's strong performance, achieving state-of-the-art FVD on the UCF101 class-conditional video generation benchmark. LARP enhances the compatibility of AR models with videos and opens up the potential to build unified high-fidelity multimodal large language models (MLLMs).

Saksham Suri, Matthew Walmer, Kamal Gupta et al.

We present a simple self-supervised method to enhance the performance of ViT features for dense downstream tasks. Our Lightweight Feature Transform (LiFT) is a straightforward and compact postprocessing network that can be applied to enhance the features of any pre-trained ViT backbone. LiFT is fast and easy to train with a self-supervised objective, and it boosts the density of ViT features for minimal extra inference cost. Furthermore, we demonstrate that LiFT can be applied with approaches that use additional task-specific downstream modules, as we integrate LiFT with ViTDet for COCO detection and segmentation. Despite the simplicity of LiFT, we find that it is not simply learning a more complex version of bilinear interpolation. Instead, our LiFT training protocol leads to several desirable emergent properties that benefit ViT features in dense downstream tasks. This includes greater scale invariance for features, and better object boundary maps. By simply training LiFT for a few epochs, we show improved performance on keypoint correspondence, detection, segmentation, and object discovery tasks. Overall, LiFT provides an easy way to unlock the benefits of denser feature arrays for a fraction of the computational cost. For more details, refer to our project page at https://www.cs.umd.edu/~sakshams/LiFT/.

Saksham Suri, Fanyi Xiao, Animesh Sinha et al.

Recently diffusion models have shown improvement in synthetic image quality as well as better control in generation. We motivate and present Gen2Det, a simple modular pipeline to create synthetic training data for object detection for free by leveraging state-of-the-art grounded image generation methods. Unlike existing works which generate individual object instances, require identifying foreground followed by pasting on other images, we simplify to directly generating scene-centric images. In addition to the synthetic data, Gen2Det also proposes a suite of techniques to best utilize the generated data, including image-level filtering, instance-level filtering, and better training recipe to account for imperfections in the generation. Using Gen2Det, we show healthy improvements on object detection and segmentation tasks under various settings and agnostic to detection methods. In the long-tailed detection setting on LVIS, Gen2Det improves the performance on rare categories by a large margin while also significantly improving the performance on other categories, e.g. we see an improvement of 2.13 Box AP and 1.84 Mask AP over just training on real data on LVIS with Mask R-CNN. In the low-data regime setting on COCO, Gen2Det consistently improves both Box and Mask AP by 2.27 and 1.85 points. In the most general detection setting, Gen2Det still demonstrates robust performance gains, e.g. it improves the Box and Mask AP on COCO by 0.45 and 0.32 points.

Sharath Girish, Tianye Li, Amrita Mazumdar et al.

Online free-viewpoint video (FVV) streaming is a challenging problem, which is relatively under-explored. It requires incremental on-the-fly updates to a volumetric representation, fast training and rendering to satisfy real-time constraints and a small memory footprint for efficient transmission. If achieved, it can enhance user experience by enabling novel applications, e.g., 3D video conferencing and live volumetric video broadcast, among others. In this work, we propose a novel framework for QUantized and Efficient ENcoding (QUEEN) for streaming FVV using 3D Gaussian Splatting (3D-GS). QUEEN directly learns Gaussian attribute residuals between consecutive frames at each time-step without imposing any structural constraints on them, allowing for high quality reconstruction and generalizability. To efficiently store the residuals, we further propose a quantization-sparsity framework, which contains a learned latent-decoder for effectively quantizing attribute residuals other than Gaussian positions and a learned gating module to sparsify position residuals. We propose to use the Gaussian viewspace gradient difference vector as a signal to separate the static and dynamic content of the scene. It acts as a guide for effective sparsity learning and speeds up training. On diverse FVV benchmarks, QUEEN outperforms the state-of-the-art online FVV methods on all metrics. Notably, for several highly dynamic scenes, it reduces the model size to just 0.7 MB per frame while training in under 5 sec and rendering at 350 FPS. Project website is at https://research.nvidia.com/labs/amri/projects/queen

Gaurav Shrivastava, Ser-Nam Lim, Abhinav Shrivastava

In this paper, we present a novel robust framework for low-level vision tasks, including denoising, object removal, frame interpolation, and super-resolution, that does not require any external training data corpus. Our proposed approach directly learns the weights of neural modules by optimizing over the corrupted test sequence, leveraging the spatio-temporal coherence and internal statistics of videos. Furthermore, we introduce a novel spatial pyramid loss that leverages the property of spatio-temporal patch recurrence in a video across the different scales of the video. This loss enhances robustness to unstructured noise in both the spatial and temporal domains. This further results in our framework being highly robust to degradation in input frames and yields state-of-the-art results on downstream tasks such as denoising, object removal, and frame interpolation. To validate the effectiveness of our approach, we conduct qualitative and quantitative evaluations on standard video datasets such as DAVIS, UCF-101, and VIMEO90K-T.

Hanyu Wang, Pengxiang Wu, Kevin Dela Rosa et al.

Image Style Transfer (IST) is an interdisciplinary topic of computer vision and art that continuously attracts researchers' interests. Different from traditional Image-guided Image Style Transfer (IIST) methods that require a style reference image as input to define the desired style, recent works start to tackle the problem in a text-guided manner, i.e., Text-guided Image Style Transfer (TIST). Compared to IIST, such approaches provide more flexibility with text-specified styles, which are useful in scenarios where the style is hard to define with reference images. Unfortunately, many TIST approaches produce undesirable artifacts in the transferred images. To address this issue, we present a novel method to achieve much improved style transfer based on text guidance. Meanwhile, to offer more flexibility than IIST and TIST, our method allows style inputs from multiple sources and modalities, enabling MultiModality-guided Image Style Transfer (MMIST). Specifically, we realize MMIST with a novel cross-modal GAN inversion method, which generates style representations consistent with specified styles. Such style representations facilitate style transfer and in principle generalize any IIST methods to MMIST. Large-scale experiments and user studies demonstrate that our method achieves state-of-the-art performance on TIST task. Furthermore, comprehensive qualitative results confirm the effectiveness of our method on MMIST task and cross-modal style interpolation.