Dongliang He

Ren Yang, Radu Timofte, Xin Li et al.

This paper reviews the Challenge on Super-Resolution of Compressed Image and Video at AIM 2022. This challenge includes two tracks. Track 1 aims at the super-resolution of compressed image, and Track~2 targets the super-resolution of compressed video. In Track 1, we use the popular dataset DIV2K as the training, validation and test sets. In Track 2, we propose the LDV 3.0 dataset, which contains 365 videos, including the LDV 2.0 dataset (335 videos) and 30 additional videos. In this challenge, there are 12 teams and 2 teams that submitted the final results to Track 1 and Track 2, respectively. The proposed methods and solutions gauge the state-of-the-art of super-resolution on compressed image and video. The proposed LDV 3.0 dataset is available at https://github.com/RenYang-home/LDV_dataset. The homepage of this challenge is at https://github.com/RenYang-home/AIM22_CompressSR.

Cong Cao, Tianwei Lin, Dongliang He et al.

A common challenge posed to robust semantic segmentation is the expensive data annotation cost. Existing semi-supervised solutions show great potential for solving this problem. Their key idea is constructing consistency regularization with unsupervised data augmentation from unlabeled data for model training. The perturbations for unlabeled data enable the consistency training loss, which benefits semi-supervised semantic segmentation. However, these perturbations destroy image context and introduce unnatural boundaries, which is harmful for semantic segmentation. Besides, the widely adopted semi-supervised learning framework, i.e. mean-teacher, suffers performance limitation since the student model finally converges to the teacher model. In this paper, first of all, we propose a context friendly differentiable geometric warping to conduct unsupervised data augmentation; secondly, a novel adversarial dual-student framework is proposed to improve the Mean-Teacher from the following two aspects: (1) dual student models are learned independently except for a stabilization constraint to encourage exploiting model diversities; (2) adversarial training scheme is applied to both students and the discriminators are resorted to distinguish reliable pseudo-label of unlabeled data for self-training. Effectiveness is validated via extensive experiments on PASCAL VOC2012 and Cityscapes. Our solution significantly improves the performance and state-of-the-art results are achieved on both datasets. Remarkably, compared with fully supervision, our solution achieves comparable mIoU of 73.4% using only 12.5% annotated data on PASCAL VOC2012. Our codes and models are available at https://github.com/cao-cong/ADS-SemiSeg.

Yili Wang, Xin Li, Kun Xu et al.

We propose a novel image retouching method by modeling the retouching process as performing a sequence of newly introduced trainable neural color operators. The neural color operator mimics the behavior of traditional color operators and learns pixelwise color transformation while its strength is controlled by a scalar. To reflect the homomorphism property of color operators, we employ equivariant mapping and adopt an encoder-decoder structure which maps the non-linear color transformation to a much simpler transformation (i.e., translation) in a high dimensional space. The scalar strength of each neural color operator is predicted using CNN based strength predictors by analyzing global image statistics. Overall, our method is rather lightweight and offers flexible controls. Experiments and user studies on public datasets show that our method consistently achieves the best results compared with SOTA methods in both quantitative measures and visual qualities. The code and pretrained models are provided at https://github.com/amberwangyili/neurop

Boyang Xia, Wenhao Wu, Haoran Wang et al. · amazon-science

It is challenging for artificial intelligence systems to achieve accurate video recognition under the scenario of low computation costs. Adaptive inference based efficient video recognition methods typically preview videos and focus on salient parts to reduce computation costs. Most existing works focus on complex networks learning with video classification based objectives. Taking all frames as positive samples, few of them pay attention to the discrimination between positive samples (salient frames) and negative samples (non-salient frames) in supervisions. To fill this gap, in this paper, we propose a novel Non-saliency Suppression Network (NSNet), which effectively suppresses the responses of non-salient frames. Specifically, on the frame level, effective pseudo labels that can distinguish between salient and non-salient frames are generated to guide the frame saliency learning. On the video level, a temporal attention module is learned under dual video-level supervisions on both the salient and the non-salient representations. Saliency measurements from both two levels are combined for exploitation of multi-granularity complementary information. Extensive experiments conducted on four well-known benchmarks verify our NSNet not only achieves the state-of-the-art accuracy-efficiency trade-off but also present a significantly faster (2.4~4.3x) practical inference speed than state-of-the-art methods. Our project page is at https://lawrencexia2008.github.io/projects/nsnet .

Zhihong Pan, Baopu Li, Dongliang He et al. · amazon-science

Deep learning based single image super-resolution models have been widely studied and superb results are achieved in upscaling low-resolution images with fixed scale factor and downscaling degradation kernel. To improve real world applicability of such models, there are growing interests to develop models optimized for arbitrary upscaling factors. Our proposed method is the first to treat arbitrary rescaling, both upscaling and downscaling, as one unified process. Using joint optimization of both directions, the proposed model is able to learn upscaling and downscaling simultaneously and achieve bidirectional arbitrary image rescaling. It improves the performance of current arbitrary upscaling models by a large margin while at the same time learns to maintain visual perception quality in downscaled images. The proposed model is further shown to be robust in cycle idempotence test, free of severe degradations in reconstruction accuracy when the downscaling-to-upscaling cycle is applied repetitively. This robustness is beneficial for image rescaling in the wild when this cycle could be applied to one image for multiple times. It also performs well on tests with arbitrary large scales and asymmetric scales, even when the model is not trained with such tasks. Extensive experiments are conducted to demonstrate the superior performance of our model.

Haoran Wang, Dongliang He, Wenhao Wu et al. · amazon-science

Image-Text Retrieval (ITR) is challenging in bridging visual and lingual modalities. Contrastive learning has been adopted by most prior arts. Except for limited amount of negative image-text pairs, the capability of constrastive learning is restricted by manually weighting negative pairs as well as unawareness of external knowledge. In this paper, we propose our novel Coupled Diversity-Sensitive Momentum Constrastive Learning (CODER) for improving cross-modal representation. Firstly, a novel diversity-sensitive contrastive learning (DCL) architecture is invented. We introduce dynamic dictionaries for both modalities to enlarge the scale of image-text pairs, and diversity-sensitiveness is achieved by adaptive negative pair weighting. Furthermore, two branches are designed in CODER. One learns instance-level embeddings from image/text, and it also generates pseudo online clustering labels for its input image/text based on their embeddings. Meanwhile, the other branch learns to query from commonsense knowledge graph to form concept-level descriptors for both modalities. Afterwards, both branches leverage DCL to align the cross-modal embedding spaces while an extra pseudo clustering label prediction loss is utilized to promote concept-level representation learning for the second branch. Extensive experiments conducted on two popular benchmarks, i.e. MSCOCO and Flicker30K, validate CODER remarkably outperforms the state-of-the-art approaches.

Zhihong Pan, Baopu Li, Dongliang He et al. · amazon-science

Great successes have been achieved using deep learning techniques for image super-resolution (SR) with fixed scales. To increase its real world applicability, numerous models have also been proposed to restore SR images with arbitrary scale factors, including asymmetric ones where images are resized to different scales along horizontal and vertical directions. Though most models are only optimized for the unidirectional upscaling task while assuming a predefined downscaling kernel for low-resolution (LR) inputs, recent models based on Invertible Neural Networks (INN) are able to increase upscaling accuracy significantly by optimizing the downscaling and upscaling cycle jointly. However, limited by the INN architecture, it is constrained to fixed integer scale factors and requires one model for each scale. Without increasing model complexity, a simple and effective invertible arbitrary rescaling network (IARN) is proposed to achieve arbitrary image rescaling by training only one model in this work. Using innovative components like position-aware scale encoding and preemptive channel splitting, the network is optimized to convert the non-invertible rescaling cycle to an effectively invertible process. It is shown to achieve a state-of-the-art (SOTA) performance in bidirectional arbitrary rescaling without compromising perceptual quality in LR outputs. It is also demonstrated to perform well on tests with asymmetric scales using the same network architecture.

Jiaxiang Tang, Kaisiyuan Wang, Hang Zhou et al.

While dynamic Neural Radiance Fields (NeRF) have shown success in high-fidelity 3D modeling of talking portraits, the slow training and inference speed severely obstruct their potential usage. In this paper, we propose an efficient NeRF-based framework that enables real-time synthesizing of talking portraits and faster convergence by leveraging the recent success of grid-based NeRF. Our key insight is to decompose the inherently high-dimensional talking portrait representation into three low-dimensional feature grids. Specifically, a Decomposed Audio-spatial Encoding Module models the dynamic head with a 3D spatial grid and a 2D audio grid. The torso is handled with another 2D grid in a lightweight Pseudo-3D Deformable Module. Both modules focus on efficiency under the premise of good rendering quality. Extensive experiments demonstrate that our method can generate realistic and audio-lips synchronized talking portrait videos, while also being highly efficient compared to previous methods.

Yuxin Song, Min Yang, Wenhao Wu et al. · amazon-science

Self-supervised video transformer pre-training has recently benefited from the mask-and-predict pipeline. They have demonstrated outstanding effectiveness on downstream video tasks and superior data efficiency on small datasets. However, temporal relation is not fully exploited by these methods. In this work, we explicitly investigate motion cues in videos as extra prediction target and propose our Masked Appearance-Motion Modeling (MAM2) framework. Specifically, we design an encoder-regressor-decoder pipeline for this task. The regressor separates feature encoding and pretext tasks completion, such that the feature extraction process is completed adequately by the encoder. In order to guide the encoder to fully excavate spatial-temporal features, two separate decoders are used for two pretext tasks of disentangled appearance and motion prediction. We explore various motion prediction targets and figure out RGB-difference is simple yet effective. As for appearance prediction, VQGAN codes are leveraged as prediction target. With our pre-training pipeline, convergence can be remarkably speed up, e.g., we only require half of epochs than state-of-the-art VideoMAE (400 v.s. 800) to achieve the competitive performance. Extensive experimental results prove that our method learns generalized video representations. Notably, our MAM2 with ViT-B achieves 82.3% on Kinects-400, 71.3% on Something-Something V2, 91.5% on UCF101, and 62.5% on HMDB51.

Hao Tang, Songhua Liu, Tianwei Lin et al.

Transformer-based models achieve favorable performance in artistic style transfer recently thanks to its global receptive field and powerful multi-head/layer attention operations. Nevertheless, the over-paramerized multi-layer structure increases parameters significantly and thus presents a heavy burden for training. Moreover, for the task of style transfer, vanilla Transformer that fuses content and style features by residual connections is prone to content-wise distortion. In this paper, we devise a novel Transformer model termed as \emph{Master} specifically for style transfer. On the one hand, in the proposed model, different Transformer layers share a common group of parameters, which (1) reduces the total number of parameters, (2) leads to more robust training convergence, and (3) is readily to control the degree of stylization via tuning the number of stacked layers freely during inference. On the other hand, different from the vanilla version, we adopt a learnable scaling operation on content features before content-style feature interaction, which better preserves the original similarity between a pair of content features while ensuring the stylization quality. We also propose a novel meta learning scheme for the proposed model so that it can not only work in the typical setting of arbitrary style transfer, but also adaptable to the few-shot setting, by only fine-tuning the Transformer encoder layer in the few-shot stage for one specific style. Text-guided few-shot style transfer is firstly achieved with the proposed framework. Extensive experiments demonstrate the superiority of Master under both zero-shot and few-shot style transfer settings.

Lin Zhang, Xin Li, Dongliang He et al.

It is widely agreed that reference-based super-resolution (RefSR) achieves superior results by referring to similar high quality images, compared to single image super-resolution (SISR). Intuitively, the more references, the better performance. However, previous RefSR methods have all focused on single-reference image training, while multiple reference images are often available in testing or practical applications. The root cause of such training-testing mismatch is the absence of publicly available multi-reference SR training datasets, which greatly hinders research efforts on multi-reference super-resolution. To this end, we construct a large-scale, multi-reference super-resolution dataset, named LMR. It contains 112,142 groups of 300x300 training images, which is 10x of the existing largest RefSR dataset. The image size is also much larger. More importantly, each group is equipped with 5 reference images with different similarity levels. Furthermore, we propose a new baseline method for multi-reference super-resolution: MRefSR, including a Multi-Reference Attention Module (MAM) for feature fusion of an arbitrary number of reference images, and a Spatial Aware Filtering Module (SAFM) for the fused feature selection. The proposed MRefSR achieves significant improvements over state-of-the-art approaches on both quantitative and qualitative evaluations. Our code and data would be made available soon.

Tianwei Lin, Honglin Lin, Fu Li et al. · amazon-science

Photo-realistic style transfer aims at migrating the artistic style from an exemplar style image to a content image, producing a result image without spatial distortions or unrealistic artifacts. Impressive results have been achieved by recent deep models. However, deep neural network based methods are too expensive to run in real-time. Meanwhile, bilateral grid based methods are much faster but still contain artifacts like overexposure. In this work, we propose the \textbf{Adaptive ColorMLP (AdaCM)}, an effective and efficient framework for universal photo-realistic style transfer. First, we find the complex non-linear color mapping between input and target domain can be efficiently modeled by a small multi-layer perceptron (ColorMLP) model. Then, in \textbf{AdaCM}, we adopt a CNN encoder to adaptively predict all parameters for the ColorMLP conditioned on each input content and style image pair. Experimental results demonstrate that AdaCM can generate vivid and high-quality stylization results. Meanwhile, our AdaCM is ultrafast and can process a 4K resolution image in 6ms on one V100 GPU.

Haoran Wang, Di Xu, Dongliang He et al.

Video-text retrieval (VTR) is an attractive yet challenging task for multi-modal understanding, which aims to search for relevant video (text) given a query (video). Existing methods typically employ completely heterogeneous visual-textual information to align video and text, whilst lacking the awareness of homogeneous high-level semantic information residing in both modalities. To fill this gap, in this work, we propose a novel visual-linguistic aligning model named HiSE for VTR, which improves the cross-modal representation by incorporating explicit high-level semantics. First, we explore the hierarchical property of explicit high-level semantics, and further decompose it into two levels, i.e. discrete semantics and holistic semantics. Specifically, for visual branch, we exploit an off-the-shelf semantic entity predictor to generate discrete high-level semantics. In parallel, a trained video captioning model is employed to output holistic high-level semantics. As for the textual modality, we parse the text into three parts including occurrence, action and entity. In particular, the occurrence corresponds to the holistic high-level semantics, meanwhile both action and entity represent the discrete ones. Then, different graph reasoning techniques are utilized to promote the interaction between holistic and discrete high-level semantics. Extensive experiments demonstrate that, with the aid of explicit high-level semantics, our method achieves the superior performance over state-of-the-art methods on three benchmark datasets, including MSR-VTT, MSVD and DiDeMo.

Yueming Lyu, Tianwei Lin, Fu Li et al.

Text-driven image manipulation remains challenging in training or inference flexibility. Conditional generative models depend heavily on expensive annotated training data. Meanwhile, recent frameworks, which leverage pre-trained vision-language models, are limited by either per text-prompt optimization or inference-time hyper-parameters tuning. In this work, we propose a novel framework named \textit{DeltaEdit} to address these problems. Our key idea is to investigate and identify a space, namely delta image and text space that has well-aligned distribution between CLIP visual feature differences of two images and CLIP textual embedding differences of source and target texts. Based on the CLIP delta space, the DeltaEdit network is designed to map the CLIP visual features differences to the editing directions of StyleGAN at training phase. Then, in inference phase, DeltaEdit predicts the StyleGAN's editing directions from the differences of the CLIP textual features. In this way, DeltaEdit is trained in a text-free manner. Once trained, it can well generalize to various text prompts for zero-shot inference without bells and whistles.

Haiyang Xu, Zhichao Zhou, Dongliang He et al.

Vision Transformer(ViT) is now dominating many vision tasks. The drawback of quadratic complexity of its token-wise multi-head self-attention (MHSA), is extensively addressed via either token sparsification or dimension reduction (in spatial or channel). However, the therein redundancy of MHSA is usually overlooked and so is the feed-forward network (FFN). To this end, we propose attention map hallucination and FFN compaction to fill in the blank. Specifically, we observe similar attention maps exist in vanilla ViT and propose to hallucinate half of the attention maps from the rest with much cheaper operations, which is called hallucinated-MHSA (hMHSA). As for FFN, we factorize its hidden-to-output projection matrix and leverage the re-parameterization technique to strengthen its capability, making it compact-FFN (cFFN). With our proposed modules, a 10$\%$-20$\%$ reduction of floating point operations (FLOPs) and parameters (Params) is achieved for various ViT-based backbones, including straight (DeiT), hybrid (NextViT) and hierarchical (PVT) structures, meanwhile, the performances are quite competitive.

Lin Zhang, Xin Li, Dongliang He et al.

Reference-based image super-resolution (RefSR) is a promising SR branch and has shown great potential in overcoming the limitations of single image super-resolution. While previous state-of-the-art RefSR methods mainly focus on improving the efficacy and robustness of reference feature transfer, it is generally overlooked that a well reconstructed SR image should enable better SR reconstruction for its similar LR images when it is referred to as. Therefore, in this work, we propose a reciprocal learning framework that can appropriately leverage such a fact to reinforce the learning of a RefSR network. Besides, we deliberately design a progressive feature alignment and selection module for further improving the RefSR task. The newly proposed module aligns reference-input images at multi-scale feature spaces and performs reference-aware feature selection in a progressive manner, thus more precise reference features can be transferred into the input features and the network capability is enhanced. Our reciprocal learning paradigm is model-agnostic and it can be applied to arbitrary RefSR models. We empirically show that multiple recent state-of-the-art RefSR models can be consistently improved with our reciprocal learning paradigm. Furthermore, our proposed model together with the reciprocal learning strategy sets new state-of-the-art performances on multiple benchmarks.

Yuxin Song, Ruolin Zhu, Min Yang et al.

Deeply learned representations have achieved superior image retrieval performance in a retrieve-then-rerank manner. Recent state-of-the-art single stage model, which heuristically fuses local and global features, achieves promising trade-off between efficiency and effectiveness. However, we notice that efficiency of existing solutions is still restricted because of their multi-scale inference paradigm. In this paper, we follow the single stage art and obtain further complexity-effectiveness balance by successfully getting rid of multi-scale testing. To achieve this goal, we abandon the widely-used convolution network giving its limitation in exploring diverse visual patterns, and resort to fully attention based framework for robust representation learning motivated by the success of Transformer. Besides applying Transformer for global feature extraction, we devise a local branch composed of window-based multi-head attention and spatial attention to fully exploit local image patterns. Furthermore, we propose to combine the hierarchical local and global features via a cross-attention module, instead of using heuristically fusion as previous art does. With our Deep Attentive Local and Global modeling framework (DALG), extensive experimental results show that efficiency can be significantly improved while maintaining competitive results with the state of the arts.

Weibo Gong, Xiaojie Jin, Xin Li et al.

Verbal videos, featuring voice-overs or text overlays, provide valuable content but present significant challenges in composition, especially when incorporating editing effects to enhance clarity and visual appeal. In this paper, we introduce the novel task of verbal video composition with editing effects. This task aims to generate coherent and visually appealing verbal videos by integrating multimodal editing effects across textual, visual, and audio categories. To achieve this, we curate a large-scale dataset of video effects compositions from publicly available sources. We then formulate this task as a generative problem, involving the identification of appropriate positions in the verbal content and the recommendation of editing effects for these positions. To address this task, we propose VCoME, a general framework that employs a large multimodal model to generate editing effects for video composition. Specifically, VCoME takes in the multimodal video context and autoregressively outputs where to apply effects within the verbal content and which effects are most appropriate for each position. VCoME also supports prompt-based control of composition density and style, providing substantial flexibility for diverse applications. Through extensive quantitative and qualitative evaluations, we clearly demonstrate the effectiveness of VCoME. A comprehensive user study shows that our method produces videos of professional quality while being 85$\times$ more efficient than professional editors.

You Li, Dewei Zhou, Fan Ma et al.

Recent Video-to-Audio (V2A) methods have achieved remarkable progress, enabling the synthesis of realistic, high-quality audio. However, they struggle with fine-grained temporal control in multi-event scenarios or when visual cues are insufficient, such as small regions, off-screen sounds, or occluded or partially visible objects. In this paper, we propose FoleyDirector, a framework that, for the first time, enables precise temporal guidance in DiT-based V2A generation while preserving the base model's audio quality and allowing seamless switching between V2A generation and temporally controlled synthesis. FoleyDirector introduces Structured Temporal Scripts (STS), a set of captions corresponding to short temporal segments, to provide richer temporal information. These features are integrated via the Script-Guided Temporal Fusion Module, which employs Temporal Script Attention to fuse STS features coherently. To handle complex multi-event scenarios, we further propose Bi-Frame Sound Synthesis, enabling parallel in-frame and out-of-frame audio generation and improving controllability. To support training and evaluation, we construct the DirectorSound dataset and introduce VGGSoundDirector and DirectorBench. Experiments demonstrate that FoleyDirector substantially enhances temporal controllability while maintaining high audio fidelity, empowering users to act as Foley directors and advancing V2A toward more expressive and controllable generation.

Haoji Zhang, Xin Gu, Jiawen Li et al.

The video reasoning ability of multimodal large language models (MLLMs) is crucial for downstream tasks like video question answering and temporal grounding. While recent approaches have explored text-based chain-of-thought (CoT) reasoning for MLLMs, these methods often suffer from limited cross-modal interaction and increased hallucination, especially with longer videos or reasoning chains. To address these challenges, we propose Video Intelligence via Tool-Augmented Learning (VITAL), a novel end-to-end agentic video reasoning framework. With a visual toolbox, the model can densely sample new video frames on demand and generate multimodal CoT for precise long video reasoning. We observe that temporal grounding and question answering are mutually beneficial for video understanding tasks. Therefore, we construct two high-quality multi-task video reasoning datasets MTVR-CoT-72k for supervised fine-tuning and MTVR-RL-110k for reinforcement learning. Moreover, we propose a Difficulty-aware Group Relative Policy Optimization algorithm (DGRPO) to mitigate difficulty imbalance in multi-task reinforcement learning. Extensive experiments on 11 challenging video understanding benchmarks demonstrate the advanced reasoning ability of VITAL, outperforming existing methods in video question answering and temporal grounding tasks, especially in long video scenarios. Code is available at https://zhang9302002.github.io/thinkingwithvideos-page/.

Wenhao Wu, Yuxiang Zhao, Yanwu Xu et al.

Long-range and short-range temporal modeling are two complementary and crucial aspects of video recognition. Most of the state-of-the-arts focus on short-range spatio-temporal modeling and then average multiple snippet-level predictions to yield the final video-level prediction. Thus, their video-level prediction does not consider spatio-temporal features of how video evolves along the temporal dimension. In this paper, we introduce a novel Dynamic Segment Aggregation (DSA) module to capture relationship among snippets. To be more specific, we attempt to generate a dynamic kernel for a convolutional operation to aggregate long-range temporal information among adjacent snippets adaptively. The DSA module is an efficient plug-and-play module and can be combined with the off-the-shelf clip-based models (i.e., TSM, I3D) to perform powerful long-range modeling with minimal overhead. The final video architecture, coined as DSANet. We conduct extensive experiments on several video recognition benchmarks (i.e., Mini-Kinetics-200, Kinetics-400, Something-Something V1 and ActivityNet) to show its superiority. Our proposed DSA module is shown to benefit various video recognition models significantly. For example, equipped with DSA modules, the top-1 accuracy of I3D ResNet-50 is improved from 74.9% to 78.2% on Kinetics-400. Codes are available at https://github.com/whwu95/DSANet.

Ren Yang, Radu Timofte, Jing Liu et al.

This paper reviews the first NTIRE challenge on quality enhancement of compressed video, with a focus on the proposed methods and results. In this challenge, the new Large-scale Diverse Video (LDV) dataset is employed. The challenge has three tracks. Tracks 1 and 2 aim at enhancing the videos compressed by HEVC at a fixed QP, while Track 3 is designed for enhancing the videos compressed by x265 at a fixed bit-rate. Besides, the quality enhancement of Tracks 1 and 3 targets at improving the fidelity (PSNR), and Track 2 targets at enhancing the perceptual quality. The three tracks totally attract 482 registrations. In the test phase, 12 teams, 8 teams and 11 teams submitted the final results of Tracks 1, 2 and 3, respectively. The proposed methods and solutions gauge the state-of-the-art of video quality enhancement. The homepage of the challenge: https://github.com/RenYang-home/NTIRE21_VEnh

Zhengyao Lv, Xiaoming Li, Xin Li et al.

Human pose transfer has received great attention due to its wide applications, yet is still a challenging task that is not well solved. Recent works have achieved great success to transfer the person image from the source to the target pose. However, most of them cannot well capture the semantic appearance, resulting in inconsistent and less realistic textures on the reconstructed results. To address this issue, we propose a new two-stage framework to handle the pose and appearance translation. In the first stage, we predict the target semantic parsing maps to eliminate the difficulties of pose transfer and further benefit the latter translation of per-region appearance style. In the second one, with the predicted target semantic maps, we suggest a new person image generation method by incorporating the region-adaptive normalization, in which it takes the per-region styles to guide the target appearance generation. Extensive experiments show that our proposed SPGNet can generate more semantic, consistent, and photo-realistic results and perform favorably against the state of the art methods in terms of quantitative and qualitative evaluation. The source code and model are available at https://github.com/cszy98/SPGNet.git.

Peihao Chen, Deng Huang, Dongliang He et al.

We study unsupervised video representation learning that seeks to learn both motion and appearance features from unlabeled video only, which can be reused for downstream tasks such as action recognition. This task, however, is extremely challenging due to 1) the highly complex spatial-temporal information in videos; and 2) the lack of labeled data for training. Unlike the representation learning for static images, it is difficult to construct a suitable self-supervised task to well model both motion and appearance features. More recently, several attempts have been made to learn video representation through video playback speed prediction. However, it is non-trivial to obtain precise speed labels for the videos. More critically, the learnt models may tend to focus on motion pattern and thus may not learn appearance features well. In this paper, we observe that the relative playback speed is more consistent with motion pattern, and thus provide more effective and stable supervision for representation learning. Therefore, we propose a new way to perceive the playback speed and exploit the relative speed between two video clips as labels. In this way, we are able to well perceive speed and learn better motion features. Moreover, to ensure the learning of appearance features, we further propose an appearance-focused task, where we enforce the model to perceive the appearance difference between two video clips. We show that optimizing the two tasks jointly consistently improves the performance on two downstream tasks, namely action recognition and video retrieval. Remarkably, for action recognition on UCF101 dataset, we achieve 93.7% accuracy without the use of labeled data for pre-training, which outperforms the ImageNet supervised pre-trained model. Code and pre-trained models can be found at https://github.com/PeihaoChen/RSPNet.

Pengcheng Yuan, Shufei Lin, Cheng Cui et al.

This paper addresses representational block named Hierarchical-Split Block, which can be taken as a plug-and-play block to upgrade existing convolutional neural networks, improves model performance significantly in a network. Hierarchical-Split Block contains many hierarchical split and concatenate connections within one single residual block. We find multi-scale features is of great importance for numerous vision tasks. Moreover, Hierarchical-Split block is very flexible and efficient, which provides a large space of potential network architectures for different applications. In this work, we present a common backbone based on Hierarchical-Split block for tasks: image classification, object detection, instance segmentation and semantic image segmentation/parsing. Our approach shows significant improvements over all these core tasks in comparison with the baseline. As shown in Figure1, for image classification, our 50-layers network(HS-ResNet50) achieves 81.28% top-1 accuracy with competitive latency on ImageNet-1k dataset. It also outperforms most state-of-the-art models. The source code and models will be available on: https://github.com/PaddlePaddle/PaddleClas

Jiazhi Guan, Zhanwang Zhang, Hang Zhou et al.

Despite recent advances in syncing lip movements with any audio waves, current methods still struggle to balance generation quality and the model's generalization ability. Previous studies either require long-term data for training or produce a similar movement pattern on all subjects with low quality. In this paper, we propose StyleSync, an effective framework that enables high-fidelity lip synchronization. We identify that a style-based generator would sufficiently enable such a charming property on both one-shot and few-shot scenarios. Specifically, we design a mask-guided spatial information encoding module that preserves the details of the given face. The mouth shapes are accurately modified by audio through modulated convolutions. Moreover, our design also enables personalized lip-sync by introducing style space and generator refinement on only limited frames. Thus the identity and talking style of a target person could be accurately preserved. Extensive experiments demonstrate the effectiveness of our method in producing high-fidelity results on a variety of scenes. Resources can be found at https://hangz-nju-cuhk.github.io/projects/StyleSync.

Zipeng Xu, Tianwei Lin, Hao Tang et al.

To achieve disentangled image manipulation, previous works depend heavily on manual annotation. Meanwhile, the available manipulations are limited to a pre-defined set the models were trained for. We propose a novel framework, i.e., Predict, Prevent, and Evaluate (PPE), for disentangled text-driven image manipulation that requires little manual annotation while being applicable to a wide variety of manipulations. Our method approaches the targets by deeply exploiting the power of the large-scale pre-trained vision-language model CLIP. Concretely, we firstly Predict the possibly entangled attributes for a given text command. Then, based on the predicted attributes, we introduce an entanglement loss to Prevent entanglements during training. Finally, we propose a new evaluation metric to Evaluate the disentangled image manipulation. We verify the effectiveness of our method on the challenging face editing task. Extensive experiments show that the proposed PPE framework achieves much better quantitative and qualitative results than the up-to-date StyleCLIP baseline.

Songhua Liu, Tianwei Lin, Dongliang He et al.

Neural painting refers to the procedure of producing a series of strokes for a given image and non-photo-realistically recreating it using neural networks. While reinforcement learning (RL) based agents can generate a stroke sequence step by step for this task, it is not easy to train a stable RL agent. On the other hand, stroke optimization methods search for a set of stroke parameters iteratively in a large search space; such low efficiency significantly limits their prevalence and practicality. Different from previous methods, in this paper, we formulate the task as a set prediction problem and propose a novel Transformer-based framework, dubbed Paint Transformer, to predict the parameters of a stroke set with a feed forward network. This way, our model can generate a set of strokes in parallel and obtain the final painting of size 512 * 512 in near real time. More importantly, since there is no dataset available for training the Paint Transformer, we devise a self-training pipeline such that it can be trained without any off-the-shelf dataset while still achieving excellent generalization capability. Experiments demonstrate that our method achieves better painting performance than previous ones with cheaper training and inference costs. Codes and models are available.

Songhua Liu, Tianwei Lin, Dongliang He et al.

Fast arbitrary neural style transfer has attracted widespread attention from academic, industrial and art communities due to its flexibility in enabling various applications. Existing solutions either attentively fuse deep style feature into deep content feature without considering feature distributions, or adaptively normalize deep content feature according to the style such that their global statistics are matched. Although effective, leaving shallow feature unexplored and without locally considering feature statistics, they are prone to unnatural output with unpleasing local distortions. To alleviate this problem, in this paper, we propose a novel attention and normalization module, named Adaptive Attention Normalization (AdaAttN), to adaptively perform attentive normalization on per-point basis. Specifically, spatial attention score is learnt from both shallow and deep features of content and style images. Then per-point weighted statistics are calculated by regarding a style feature point as a distribution of attention-weighted output of all style feature points. Finally, the content feature is normalized so that they demonstrate the same local feature statistics as the calculated per-point weighted style feature statistics. Besides, a novel local feature loss is derived based on AdaAttN to enhance local visual quality. We also extend AdaAttN to be ready for video style transfer with slight modifications. Experiments demonstrate that our method achieves state-of-the-art arbitrary image/video style transfer. Codes and models are available.

Min Yang, Dongliang He, Miao Fan et al.

Image Retrieval is a fundamental task of obtaining images similar to the query one from a database. A common image retrieval practice is to firstly retrieve candidate images via similarity search using global image features and then re-rank the candidates by leveraging their local features. Previous learning-based studies mainly focus on either global or local image representation learning to tackle the retrieval task. In this paper, we abandon the two-stage paradigm and seek to design an effective single-stage solution by integrating local and global information inside images into compact image representations. Specifically, we propose a Deep Orthogonal Local and Global (DOLG) information fusion framework for end-to-end image retrieval. It attentively extracts representative local information with multi-atrous convolutions and self-attention at first. Components orthogonal to the global image representation are then extracted from the local information. At last, the orthogonal components are concatenated with the global representation as a complementary, and then aggregation is performed to generate the final representation. The whole framework is end-to-end differentiable and can be trained with image-level labels. Extensive experimental results validate the effectiveness of our solution and show that our model achieves state-of-the-art image retrieval performances on Revisited Oxford and Paris datasets.

Hengyuan Zhao, Wenhao Wu, Yihao Liu et al.

In this paper, we present a fast exemplar-based image colorization approach using color embeddings named Color2Embed. Generally, due to the difficulty of obtaining input and ground truth image pairs, it is hard to train a exemplar-based colorization model with unsupervised and unpaired training manner. Current algorithms usually strive to achieve two procedures: i) retrieving a large number of reference images with high similarity for preparing training dataset, which is inevitably time-consuming and tedious; ii) designing complicated modules to transfer the colors of the reference image to the target image, by calculating and leveraging the deep semantic correspondence between them (e.g., non-local operation), which is computationally expensive during testing. Contrary to the previous methods, we adopt a self-augmented self-reference learning scheme, where the reference image is generated by graphical transformations from the original colorful one whereby the training can be formulated in a paired manner. Second, in order to reduce the process time, our method explicitly extracts the color embeddings and exploits a progressive style feature Transformation network, which injects the color embeddings into the reconstruction of the final image. Such design is much more lightweight and intelligible, achieving appealing performance with fast processing speed.

Deng Huang, Wenhao Wu, Weiwen Hu et al.

We study self-supervised video representation learning, which is a challenging task due to 1) lack of labels for explicit supervision; 2) unstructured and noisy visual information. Existing methods mainly use contrastive loss with video clips as the instances and learn visual representation by discriminating instances from each other, but they need a careful treatment of negative pairs by either relying on large batch sizes, memory banks, extra modalities or customized mining strategies, which inevitably includes noisy data. In this paper, we observe that the consistency between positive samples is the key to learn robust video representation. Specifically, we propose two tasks to learn the appearance and speed consistency, respectively. The appearance consistency task aims to maximize the similarity between two clips of the same video with different playback speeds. The speed consistency task aims to maximize the similarity between two clips with the same playback speed but different appearance information. We show that optimizing the two tasks jointly consistently improves the performance on downstream tasks, e.g., action recognition and video retrieval. Remarkably, for action recognition on the UCF-101 dataset, we achieve 90.8\% accuracy without using any extra modalities or negative pairs for unsupervised pretraining, which outperforms the ImageNet supervised pretrained model. Codes and models will be available.

Manyu Zhu, Dongliang He, Xin Li et al.

Inpainting arbitrary missing regions is challenging because learning valid features for various masked regions is nontrivial. Though U-shaped encoder-decoder frameworks have been witnessed to be successful, most of them share a common drawback of mask unawareness in feature extraction because all convolution windows (or regions), including those with various shapes of missing pixels, are treated equally and filtered with fixed learned kernels. To this end, we propose our novel mask-aware inpainting solution. Firstly, a Mask-Aware Dynamic Filtering (MADF) module is designed to effectively learn multi-scale features for missing regions in the encoding phase. Specifically, filters for each convolution window are generated from features of the corresponding region of the mask. The second fold of mask awareness is achieved by adopting Point-wise Normalization (PN) in our decoding phase, considering that statistical natures of features at masked points differentiate from those of unmasked points. The proposed PN can tackle this issue by dynamically assigning point-wise scaling factor and bias. Lastly, our model is designed to be an end-to-end cascaded refinement one. Supervision information such as reconstruction loss, perceptual loss and total variation loss is incrementally leveraged to boost the inpainting results from coarse to fine. Effectiveness of the proposed framework is validated both quantitatively and qualitatively via extensive experiments on three public datasets including Places2, CelebA and Paris StreetView.

Tianwei Lin, Zhuoqi Ma, Fu Li et al.

Artistic style transfer aims at migrating the style from an example image to a content image. Currently, optimization-based methods have achieved great stylization quality, but expensive time cost restricts their practical applications. Meanwhile, feed-forward methods still fail to synthesize complex style, especially when holistic global and local patterns exist. Inspired by the common painting process of drawing a draft and revising the details, we introduce a novel feed-forward method named Laplacian Pyramid Network (LapStyle). LapStyle first transfers global style patterns in low-resolution via a Drafting Network. It then revises the local details in high-resolution via a Revision Network, which hallucinates a residual image according to the draft and the image textures extracted by Laplacian filtering. Higher resolution details can be easily generated by stacking Revision Networks with multiple Laplacian pyramid levels. The final stylized image is obtained by aggregating outputs of all pyramid levels. %We also introduce a patch discriminator to better learn local patterns adversarially. Experiments demonstrate that our method can synthesize high quality stylized images in real time, where holistic style patterns are properly transferred.

Cheng Cui, Ruoyu Guo, Yuning Du et al.

Recently, research efforts have been concentrated on revealing how pre-trained model makes a difference in neural network performance. Self-supervision and semi-supervised learning technologies have been extensively explored by the community and are proven to be of great potential in obtaining a powerful pre-trained model. However, these models require huge training costs (i.e., hundreds of millions of images or training iterations). In this paper, we propose to improve existing baseline networks via knowledge distillation from off-the-shelf pre-trained big powerful models. Different from existing knowledge distillation frameworks which require student model to be consistent with both soft-label generated by teacher model and hard-label annotated by humans, our solution performs distillation by only driving prediction of the student model consistent with that of the teacher model. Therefore, our distillation setting can get rid of manually labeled data and can be trained with extra unlabeled data to fully exploit capability of teacher model for better learning. We empirically find that such simple distillation settings perform extremely effective, for example, the top-1 accuracy on ImageNet-1k validation set of MobileNetV3-large and ResNet50-D can be significantly improved from 75.2% to 79% and 79.1% to 83%, respectively. We have also thoroughly analyzed what are dominant factors that affect the distillation performance and how they make a difference. Extensive downstream computer vision tasks, including transfer learning, object detection and semantic segmentation, can significantly benefit from the distilled pretrained models. All our experiments are implemented based on PaddlePaddle, codes and a series of improved pretrained models with ssld suffix are available in PaddleClas.

Wenhao Wu, Dongliang He, Tianwei Lin et al.

Conventionally, spatiotemporal modeling network and its complexity are the two most concentrated research topics in video action recognition. Existing state-of-the-art methods have achieved excellent accuracy regardless of the complexity meanwhile efficient spatiotemporal modeling solutions are slightly inferior in performance. In this paper, we attempt to acquire both efficiency and effectiveness simultaneously. First of all, besides traditionally treating H x W x T video frames as space-time signal (viewing from the Height-Width spatial plane), we propose to also model video from the other two Height-Time and Width-Time planes, to capture the dynamics of video thoroughly. Secondly, our model is designed based on 2D CNN backbones and model complexity is well kept in mind by design. Specifically, we introduce a novel multi-view fusion (MVF) module to exploit video dynamics using separable convolution for efficiency. It is a plug-and-play module and can be inserted into off-the-shelf 2D CNNs to form a simple yet effective model called MVFNet. Moreover, MVFNet can be thought of as a generalized video modeling framework and it can specialize to be existing methods such as C2D, SlowOnly, and TSM under different settings. Extensive experiments are conducted on popular benchmarks (i.e., Something-Something V1 & V2, Kinetics, UCF-101, and HMDB-51) to show its superiority. The proposed MVFNet can achieve state-of-the-art performance with 2D CNN's complexity.

Dario Fuoli, Zhiwu Huang, Martin Danelljan et al.

This paper reviews the NTIRE 2020 challenge on video quality mapping (VQM), which addresses the issues of quality mapping from source video domain to target video domain. The challenge includes both a supervised track (track 1) and a weakly-supervised track (track 2) for two benchmark datasets. In particular, track 1 offers a new Internet video benchmark, requiring algorithms to learn the map from more compressed videos to less compressed videos in a supervised training manner. In track 2, algorithms are required to learn the quality mapping from one device to another when their quality varies substantially and weakly-aligned video pairs are available. For track 1, in total 7 teams competed in the final test phase, demonstrating novel and effective solutions to the problem. For track 2, some existing methods are evaluated, showing promising solutions to the weakly-supervised video quality mapping problem.

Kai Zhang, Shuhang Gu, Radu Timofte et al.

This paper reviews the NTIRE 2020 challenge on perceptual extreme super-resolution with focus on proposed solutions and results. The challenge task was to super-resolve an input image with a magnification factor 16 based on a set of prior examples of low and corresponding high resolution images. The goal is to obtain a network design capable to produce high resolution results with the best perceptual quality and similar to the ground truth. The track had 280 registered participants, and 19 teams submitted the final results. They gauge the state-of-the-art in single image super-resolution.

Wenhao Wu, Dongliang He, Xiao Tan et al.

Though action recognition in videos has achieved great success recently, it remains a challenging task due to the massive computational cost. Designing lightweight networks is a possible solution, but it may degrade the recognition performance. In this paper, we innovatively propose a general dynamic inference idea to improve inference efficiency by leveraging the variation in the distinguishability of different videos. The dynamic inference approach can be achieved from aspects of the network depth and the number of input video frames, or even in a joint input-wise and network depth-wise manner. In a nutshell, we treat input frames and network depth of the computational graph as a 2-dimensional grid, and several checkpoints are placed on this grid in advance with a prediction module. The inference is carried out progressively on the grid by following some predefined route, whenever the inference process comes across a checkpoint, an early prediction can be made depending on whether the early stop criteria meets. For the proof-of-concept purpose, we instantiate three dynamic inference frameworks using two well-known backbone CNNs. In these instances, we overcome the drawback of limited temporal coverage resulted from an early prediction by a novel frame permutation scheme, and alleviate the conflict between progressive computation and video temporal relation modeling by introducing an online temporal shift module. Extensive experiments are conducted to thoroughly analyze the effectiveness of our ideas and to inspire future research efforts. Results on various datasets also evident the superiority of our approach.

Ya Wang, Dongliang He, Fu Li et al.

Images or videos always contain multiple objects or actions. Multi-label recognition has been witnessed to achieve pretty performance attribute to the rapid development of deep learning technologies. Recently, graph convolution network (GCN) is leveraged to boost the performance of multi-label recognition. However, what is the best way for label correlation modeling and how feature learning can be improved with label system awareness are still unclear. In this paper, we propose a label graph superimposing framework to improve the conventional GCN+CNN framework developed for multi-label recognition in the following two aspects. Firstly, we model the label correlations by superimposing label graph built from statistical co-occurrence information into the graph constructed from knowledge priors of labels, and then multi-layer graph convolutions are applied on the final superimposed graph for label embedding abstraction. Secondly, we propose to leverage embedding of the whole label system for better representation learning. In detail, lateral connections between GCN and CNN are added at shallow, middle and deep layers to inject information of label system into backbone CNN for label-awareness in the feature learning process. Extensive experiments are carried out on MS-COCO and Charades datasets, showing that our proposed solution can greatly improve the recognition performance and achieves new state-of-the-art recognition performance.

Fan Yang, Xiao Liu, Dongliang He et al.

In this work, we introduce a new problem, named as {\em story-preserving long video truncation}, that requires an algorithm to automatically truncate a long-duration video into multiple short and attractive sub-videos with each one containing an unbroken story. This differs from traditional video highlight detection or video summarization problems in that each sub-video is required to maintain a coherent and integral story, which is becoming particularly important for resource-production video sharing platforms such as Youtube, Facebook, TikTok, Kwai, etc. To address the problem, we collect and annotate a new large video truncation dataset, named as TruNet, which contains 1470 videos with on average 11 short stories per video. With the new dataset, we further develop and train a neural architecture for video truncation that consists of two components: a Boundary Aware Network (BAN) and a Fast-Forward Long Short-Term Memory (FF-LSTM). We first use the BAN to generate high quality temporal proposals by jointly considering frame-level attractiveness and boundaryness. We then apply the FF-LSTM, which tends to capture high-order dependencies among a sequence of frames, to decide whether a temporal proposal is a coherent and integral story. We show that our proposed framework outperforms existing approaches for the story-preserving long video truncation problem in both quantitative measures and user-study. The dataset is available for public academic research usage at https://ai.baidu.com/broad/download.

Xin Li, Tianwei Lin, Xiao Liu et al.

Existing action localization approaches adopt shallow temporal convolutional networks (\ie, TCN) on 1D feature map extracted from video frames. In this paper, we empirically find that stacking more conventional temporal convolution layers actually deteriorates action classification performance, possibly ascribing to that all channels of 1D feature map, which generally are highly abstract and can be regarded as latent concepts, are excessively recombined in temporal convolution. To address this issue, we introduce a novel concept-wise temporal convolution (CTC) layer as an alternative to conventional temporal convolution layer for training deeper action localization networks. Instead of recombining latent concepts, CTC layer deploys a number of temporal filters to each concept separately with shared filter parameters across concepts. Thus can capture common temporal patterns of different concepts and significantly enrich representation ability. Via stacking CTC layers, we proposed a deep concept-wise temporal convolutional network (C-TCN), which boosts the state-of-the-art action localization performance on THUMOS'14 from 42.8 to 52.1 in terms of mAP(\%), achieving a relative improvement of 21.7\%. Favorable result is also obtained on ActivityNet.

Wenhao Wu, Dongliang He, Xiao Tan et al.

Video Recognition has drawn great research interest and great progress has been made. A suitable frame sampling strategy can improve the accuracy and efficiency of recognition. However, mainstream solutions generally adopt hand-crafted frame sampling strategies for recognition. It could degrade the performance, especially in untrimmed videos, due to the variation of frame-level saliency. To this end, we concentrate on improving untrimmed video classification via developing a learning-based frame sampling strategy. We intuitively formulate the frame sampling procedure as multiple parallel Markov decision processes, each of which aims at picking out a frame/clip by gradually adjusting an initial sampling. Then we propose to solve the problems with multi-agent reinforcement learning (MARL). Our MARL framework is composed of a novel RNN-based context-aware observation network which jointly models context information among nearby agents and historical states of a specific agent, a policy network which generates the probability distribution over a predefined action space at each step and a classification network for reward calculation as well as final recognition. Extensive experimental results show that our MARL-based scheme remarkably outperforms hand-crafted strategies with various 2D and 3D baseline methods. Our single RGB model achieves a comparable performance of ActivityNet v1.3 champion submission with multi-modal multi-model fusion and new state-of-the-art results on YouTube Birds and YouTube Cars.

Chao Li, Dongliang He, Xiao Liu et al.

Recently, image super-resolution has been widely studied and achieved significant progress by leveraging the power of deep convolutional neural networks. However, there has been limited advancement in video super-resolution (VSR) due to the complex temporal patterns in videos. In this paper, we investigate how to adapt state-of-the-art methods of image super-resolution for video super-resolution. The proposed adapting method is straightforward. The information among successive frames is well exploited, while the overhead on the original image super-resolution method is negligible. Furthermore, we propose a learning-based method to ensemble the outputs from multiple super-resolution models. Our methods show superior performance and rank second in the NTIRE2019 Video Super-Resolution Challenge Track 1.

Dongliang He, Xiang Zhao, Jizhou Huang et al.

The task of video grounding, which temporally localizes a natural language description in a video, plays an important role in understanding videos. Existing studies have adopted strategies of sliding window over the entire video or exhaustively ranking all possible clip-sentence pairs in a pre-segmented video, which inevitably suffer from exhaustively enumerated candidates. To alleviate this problem, we formulate this task as a problem of sequential decision making by learning an agent which regulates the temporal grounding boundaries progressively based on its policy. Specifically, we propose a reinforcement learning based framework improved by multi-task learning and it shows steady performance gains by considering additional supervised boundary information during training. Our proposed framework achieves state-of-the-art performance on ActivityNet'18 DenseCaption dataset and Charades-STA dataset while observing only 10 or less clips per video.

Dongliang He, Zhichao Zhou, Chuang Gan et al.

Despite the success of deep learning for static image understanding, it remains unclear what are the most effective network architectures for the spatial-temporal modeling in videos. In this paper, in contrast to the existing CNN+RNN or pure 3D convolution based approaches, we explore a novel spatial temporal network (StNet) architecture for both local and global spatial-temporal modeling in videos. Particularly, StNet stacks N successive video frames into a \emph{super-image} which has 3N channels and applies 2D convolution on super-images to capture local spatial-temporal relationship. To model global spatial-temporal relationship, we apply temporal convolution on the local spatial-temporal feature maps. Specifically, a novel temporal Xception block is proposed in StNet. It employs a separate channel-wise and temporal-wise convolution over the feature sequence of video. Extensive experiments on the Kinetics dataset demonstrate that our framework outperforms several state-of-the-art approaches in action recognition and can strike a satisfying trade-off between recognition accuracy and model complexity. We further demonstrate the generalization performance of the leaned video representations on the UCF101 dataset.

Dongliang He, Fu Li, Qijie Zhao et al.

In this report, our approach to tackling the task of ActivityNet 2018 Kinetics-600 challenge is described in detail. Though spatial-temporal modelling methods, which adopt either such end-to-end framework as I3D \cite{i3d} or two-stage frameworks (i.e., CNN+RNN), have been proposed in existing state-of-the-arts for this task, video modelling is far from being well solved. In this challenge, we propose spatial-temporal network (StNet) for better joint spatial-temporal modelling and comprehensively video understanding. Besides, given that multi-modal information is contained in video source, we manage to integrate both early-fusion and later-fusion strategy of multi-modal information via our proposed improved temporal Xception network (iTXN) for video understanding. Our StNet RGB single model achieves 78.99\% top-1 precision in the Kinetics-600 validation set and that of our improved temporal Xception network which integrates RGB, flow and audio modalities is up to 82.35\%. After model ensemble, we achieve top-1 precision as high as 85.0\% on the validation set and rank No.1 among all submissions.