Abstract:The field of video generation has made remarkable advancements, yet there remains a pressing need for a clear, systematic recipe that can guide the development of robust and scalable models. In this work, we present a comprehensive study that systematically explores the interplay of model architectures, training recipes, and data curation strategies, culminating in a simple and scalable text-image-conditioned video generation method, named STIV. Our framework integrates image condition into a Diffusion Transformer (DiT) through frame replacement, while incorporating text conditioning via a joint image-text conditional classifier-free guidance. This design enables STIV to perform both text-to-video (T2V) and text-image-to-video (TI2V) tasks simultaneously. Additionally, STIV can be easily extended to various applications, such as video prediction, frame interpolation, multi-view generation, and long video generation, etc. With comprehensive ablation studies on T2I, T2V, and TI2V, STIV demonstrate strong performance, despite its simple design. An 8.7B model with 512 resolution achieves 83.1 on VBench T2V, surpassing both leading open and closed-source models like CogVideoX-5B, Pika, Kling, and Gen-3. The same-sized model also achieves a state-of-the-art result of 90.1 on VBench I2V task at 512 resolution. By providing a transparent and extensible recipe for building cutting-edge video generation models, we aim to empower future research and accelerate progress toward more versatile and reliable video generation solutions.
Abstract:Recent advancements in multimodal models highlight the value of rewritten captions for improving performance, yet key challenges remain. For example, while synthetic captions often provide superior quality and image-text alignment, it is not clear whether they can fully replace AltTexts: the role of synthetic captions and their interaction with original web-crawled AltTexts in pre-training is still not well understood. Moreover, different multimodal foundation models may have unique preferences for specific caption formats, but efforts to identify the optimal captions for each model remain limited. In this work, we propose a novel, controllable, and scalable captioning pipeline designed to generate diverse caption formats tailored to various multimodal models. By examining Short Synthetic Captions (SSC) towards Dense Synthetic Captions (DSC+) as case studies, we systematically explore their effects and interactions with AltTexts across models such as CLIP, multimodal LLMs, and diffusion models. Our findings reveal that a hybrid approach that keeps both synthetic captions and AltTexts can outperform the use of synthetic captions alone, improving both alignment and performance, with each model demonstrating preferences for particular caption formats. This comprehensive analysis provides valuable insights into optimizing captioning strategies, thereby advancing the pre-training of multimodal foundation models.
Abstract:Instruction-based image editing improves the controllability and flexibility of image manipulation via natural commands without elaborate descriptions or regional masks. However, human instructions are sometimes too brief for current methods to capture and follow. Multimodal large language models (MLLMs) show promising capabilities in cross-modal understanding and visual-aware response generation via LMs. We investigate how MLLMs facilitate edit instructions and present MLLM-Guided Image Editing (MGIE). MGIE learns to derive expressive instructions and provides explicit guidance. The editing model jointly captures this visual imagination and performs manipulation through end-to-end training. We evaluate various aspects of Photoshop-style modification, global photo optimization, and local editing. Extensive experimental results demonstrate that expressive instructions are crucial to instruction-based image editing, and our MGIE can lead to a notable improvement in automatic metrics and human evaluation while maintaining competitive inference efficiency.
Abstract:Over the past few years, developing a broad, universal, and general-purpose computer vision system has become a hot topic. A powerful universal system would be capable of solving diverse vision tasks simultaneously without being restricted to a specific problem or a specific data domain, which is of great importance in practical real-world computer vision applications. This study pushes the direction forward by concentrating on the million-scale multi-domain universal object detection problem. The problem is not trivial due to its complicated nature in terms of cross-dataset category label duplication, label conflicts, and the hierarchical taxonomy handling. Moreover, what is the resource-efficient way to utilize emerging large pre-trained vision models for million-scale cross-dataset object detection remains an open challenge. This paper tries to address these challenges by introducing our practices in label handling, hierarchy-aware loss design and resource-efficient model training with a pre-trained large model. Our method is ranked second in the object detection track of Robust Vision Challenge 2022 (RVC 2022). We hope our detailed study would serve as an alternative practice paradigm for similar problems in the community. The code is available at https://github.com/linfeng93/Large-UniDet.
Abstract:With the wide and deep adoption of deep learning models in real applications, there is an increasing need to model and learn the representations of the neural networks themselves. These models can be used to estimate attributes of different neural network architectures such as the accuracy and latency, without running the actual training or inference tasks. In this paper, we propose a neural architecture representation model that can be used to estimate these attributes holistically. Specifically, we first propose a simple and effective tokenizer to encode both the operation and topology information of a neural network into a single sequence. Then, we design a multi-stage fusion transformer to build a compact vector representation from the converted sequence. For efficient model training, we further propose an information flow consistency augmentation and correspondingly design an architecture consistency loss, which brings more benefits with less augmentation samples compared with previous random augmentation strategies. Experiment results on NAS-Bench-101, NAS-Bench-201, DARTS search space and NNLQP show that our proposed framework can be used to predict the aforementioned latency and accuracy attributes of both cell architectures and whole deep neural networks, and achieves promising performance.
Abstract:Since the vision transformer (ViT) has achieved impressive performance in image classification, an increasing number of researchers pay their attentions to designing more efficient vision transformer models. A general research line is reducing computational cost of self attention modules by adopting sparse attention or using local attention windows. In contrast, we propose to design high performance transformer based architectures by densifying the attention pattern. Specifically, we propose cross attention among blocks of ViT (CabViT), which uses tokens from previous blocks in the same stage as extra input to the multi-head attention of transformers. The proposed CabViT enhances the interactions of tokens across blocks with potentially different semantics, and encourages more information flows to the lower levels, which together improves model performance and model convergence with limited extra cost. Based on the proposed CabViT, we design a series of CabViT models which achieve the best trade-off between model size, computational cost and accuracy. For instance without the need of knowledge distillation to strength the training, CabViT achieves 83.0% top-1 accuracy on Imagenet with only 16.3 million parameters and about 3.9G FLOPs, saving almost half parameters and 13% computational cost while gaining 0.9% higher accuracy compared with ConvNext, use 52% of parameters but gaining 0.6% accuracy compared with distilled EfficientFormer
Abstract:Transformers have achieved tremendous success in various computer vision tasks. By borrowing design concepts from transformers, many studies revolutionized CNNs and showed remarkable results. This paper falls in this line of studies. More specifically, we introduce a convolutional neural network architecture named ParCNetV2, which extends position-aware circular convolution (ParCNet) with oversized convolutions and strengthens attention through bifurcate gate units. The oversized convolution utilizes a kernel with $2\times$ the input size to model long-range dependencies through a global receptive field. Simultaneously, it achieves implicit positional encoding by removing the shift-invariant property from convolutional kernels, i.e., the effective kernels at different spatial locations are different when the kernel size is twice as large as the input size. The bifurcate gate unit implements an attention mechanism similar to self-attention in transformers. It splits the input into two branches, one serves as feature transformation while the other serves as attention weights. The attention is applied through element-wise multiplication of the two branches. Besides, we introduce a unified local-global convolution block to unify the design of the early and late stage convolutional blocks. Extensive experiments demonstrate that our method outperforms other pure convolutional neural networks as well as neural networks hybridizing CNNs and transformers.
Abstract:Transformer models have made tremendous progress in various fields in recent years. In the field of computer vision, vision transformers (ViTs) also become strong alternatives to convolutional neural networks (ConvNets), yet they have not been able to replace ConvNets since both have their own merits. For instance, ViTs are good at extracting global features with attention mechanisms while ConvNets are more efficient in modeling local relationships due to their strong inductive bias. A natural idea that arises is to combine the strengths of both ConvNets and ViTs to design new structures. In this paper, we propose a new basic neural network operator named position-aware circular convolution (ParC) and its accelerated version Fast-ParC. The ParC operator can capture global features by using a global kernel and circular convolution while keeping location sensitiveness by employing position embeddings. Our Fast-ParC further reduces the O(n2) time complexity of ParC to O(n log n) using Fast Fourier Transform. This acceleration makes it possible to use global convolution in the early stages of models with large feature maps, yet still maintains the overall computational cost comparable with using 3x3 or 7x7 kernels. The proposed operation can be used in a plug-and-play manner to 1) convert ViTs to pure-ConvNet architecture to enjoy wider hardware support and achieve higher inference speed; 2) replacing traditional convolutions in the deep stage of ConvNets to improve accuracy by enlarging the effective receptive field. Experiment results show that our ParC op can effectively enlarge the receptive field of traditional ConvNets, and adopting the proposed op benefits both ViTs and ConvNet models on all three popular vision tasks, image classification, object
Abstract:Active learning is an important technology for automated machine learning systems. In contrast to Neural Architecture Search (NAS) which aims at automating neural network architecture design, active learning aims at automating training data selection. It is especially critical for training a long-tailed task, in which positive samples are sparsely distributed. Active learning alleviates the expensive data annotation issue through incrementally training models powered with efficient data selection. Instead of annotating all unlabeled samples, it iteratively selects and annotates the most valuable samples. Active learning has been popular in image classification, but has not been fully explored in object detection. Most of current approaches on object detection are evaluated with different settings, making it difficult to fairly compare their performance. To facilitate the research in this field, this paper contributes an active learning benchmark framework named as ALBench for evaluating active learning in object detection. Developed on an automatic deep model training system, this ALBench framework is easy-to-use, compatible with different active learning algorithms, and ensures the same training and testing protocols. We hope this automated benchmark system help researchers to easily reproduce literature's performance and have objective comparisons with prior arts. The code will be release through Github.
Abstract:Automated machine learning systems for non-experts could be critical for industries to adopt artificial intelligence to their own applications. This paper detailed the engineering system implementation of an automated machine learning system called YMIR, which completely relies on graphical interface to interact with users. After importing training/validation data into the system, a user without AI knowledge can label the data, train models, perform data mining and evaluation by simply clicking buttons. The paper described: 1) Open implementation of model training and inference through docker containers. 2) Implementation of task and resource management. 3) Integration of Labeling software. 4) Implementation of HCI (Human Computer Interaction) with a rebuilt collaborative development paradigm. We also provide subsequent case study on training models with the system. We hope this paper can facilitate the prosperity of our automated machine learning community from industry application perspective. The code of the system has already been released to GitHub (https://github.com/industryessentials/ymir).