The objective of this work is to explore how to effectively and efficiently adapt pre-trained foundation models to various downstream tasks of image semantic segmentation. Conventional methods usually fine-tuned the whole networks for each specific dataset and it was burdensome to store the massive parameters of these networks. A few recent works attempted to insert some trainable parameters into the frozen network to learn visual prompts for efficient tuning. However, these works significantly modified the original structure of standard modules, making them inoperable on many existing high-speed inference devices, where standard modules and their parameters have been embedded. To facilitate prompt-based semantic segmentation, we propose a novel Inter-Stage Prompt-Matched Framework, which maintains the original structure of the foundation model while generating visual prompts adaptively for task-oriented tuning. Specifically, the pre-trained model is first divided into multiple stages, and their parameters are frozen and shared for all semantic segmentation tasks. A lightweight module termed Semantic-aware Prompt Matcher is then introduced to hierarchically interpolate between two stages to learn reasonable prompts for each specific task under the guidance of interim semantic maps. In this way, we can better stimulate the pre-trained knowledge of the frozen model to learn semantic concepts effectively on downstream datasets. Extensive experiments conducted on five benchmarks show that the proposed method can achieve a promising trade-off between parameter efficiency and performance effectiveness.
In this paper, we present an end-to-end approach to generate high-resolution person images conditioned on texts only. State-of-the-art text-to-image generation models are mainly designed for center-object generation, e.g., flowers and birds. Unlike center-placed objects with similar shapes and orientation, person image generation is a more challenging task, for which we observe the followings: 1) the generated images for the same person exhibit visual details with identity-consistency, e.g., identity-related textures/clothes/shoes across the images, and 2) those images should be discriminant for being robust against the inter-person variations caused by visual ambiguities. To address the above challenges, we develop an effective generative model to produce person images with two novel mechanisms. In particular, our first mechanism (called T-Person-GAN-ID) is to integrate the one-stream generator with an identity-preserving network such that the representations of generated data are regularized in their feature space to ensure the identity-consistency. The second mechanism (called T-Person-GAN-ID-MM) is based on the manifold mix-up to produce mixed images via the linear interpolation across generated images from different manifold identities, and we further enforce such interpolated images to be linearly classified in the feature space. This amounts to learning a linear classification boundary that can perfectly separate images from two identities. Our proposed method is empirically validated to achieve a remarkable improvement in text-to-person image generation. Our architecture is orthogonal to StackGAN++ , and focuses on person image generation, with all of them together to enrich the spectrum of GANs for the image generation task. Codes are available on \url{https://github.com/linwu-github/Person-Image-Generation.git}.
In computer vision, fine-tuning is the de-facto approach to leverage pre-trained vision models to perform downstream tasks. However, deploying it in practice is quite challenging, due to adopting parameter inefficient global update and heavily relying on high-quality downstream data. Recently, prompt-based learning, which adds a task-relevant prompt to adapt the downstream tasks to pre-trained models, has drastically boosted the performance of many natural language downstream tasks. In this work, we extend this notable transfer ability benefited from prompt into vision models as an alternative to fine-tuning. To this end, we propose parameter-efficient Prompt tuning (Pro-tuning) to adapt frozen vision models to various downstream vision tasks. The key to Pro-tuning is prompt-based tuning, i.e., learning task-specific vision prompts for downstream input images with the pre-trained model frozen. By only training a few additional parameters, it can work on diverse CNN-based and Transformer-based architectures. Extensive experiments evidence that Pro-tuning outperforms fine-tuning in a broad range of vision tasks and scenarios, including image classification (generic objects, class imbalance, image corruption, adversarial robustness, and out-of-distribution generalization), and dense prediction tasks such as object detection and semantic segmentation.
Large-scale vision-language pre-training has shown impressive advances in a wide range of downstream tasks. Existing methods mainly model the cross-modal alignment by the similarity of the global representations of images and texts, or advanced cross-modal attention upon image and text features. However, they fail to explicitly learn the fine-grained semantic alignment between visual regions and textual phrases, as only global image-text alignment information is available. In this paper, we introduce LOUPE, a fine-grained semantically aLigned visiOn-langUage PrE-training framework, which learns fine-grained semantic alignment from the novel perspective of game-theoretic interactions. To efficiently compute the game-theoretic interactions, we further propose an uncertainty-aware neural Shapley interaction learning module. Experiments show that LOUPE achieves state-of-the-art on image-text retrieval benchmarks. Without any object-level human annotations and fine-tuning, LOUPE achieves competitive performance on object detection and visual grounding. More importantly, LOUPE opens a new promising direction of learning fine-grained semantics from large-scale raw image-text pairs.
Understanding human emotions is a crucial ability for intelligent robots to provide better human-robot interactions. The existing works are limited to trimmed video-level emotion classification, failing to locate the temporal window corresponding to the emotion. In this paper, we introduce a new task, named Temporal Emotion Localization in videos~(TEL), which aims to detect human emotions and localize their corresponding temporal boundaries in untrimmed videos with aligned subtitles. TEL presents three unique challenges compared to temporal action localization: 1) The emotions have extremely varied temporal dynamics; 2) The emotion cues are embedded in both appearances and complex plots; 3) The fine-grained temporal annotations are complicated and labor-intensive. To address the first two challenges, we propose a novel dilated context integrated network with a coarse-fine two-stream architecture. The coarse stream captures varied temporal dynamics by modeling multi-granularity temporal contexts. The fine stream achieves complex plots understanding by reasoning the dependency between the multi-granularity temporal contexts from the coarse stream and adaptively integrates them into fine-grained video segment features. To address the third challenge, we introduce a cross-modal consensus learning paradigm, which leverages the inherent semantic consensus between the aligned video and subtitle to achieve weakly-supervised learning. We contribute a new testing set with 3,000 manually-annotated temporal boundaries so that future research on the TEL problem can be quantitatively evaluated. Extensive experiments show the effectiveness of our approach on temporal emotion localization. The repository of this work is at https://github.com/YYJMJC/Temporal-Emotion-Localization-in-Videos.
With the development of generative-based self-supervised learning (SSL) approaches like BeiT and MAE, how to learn good representations by masking random patches of the input image and reconstructing the missing information has grown in concern. However, BeiT and PeCo need a "pre-pretraining" stage to produce discrete codebooks for masked patches representing. MAE does not require a pre-training codebook process, but setting pixels as reconstruction targets may introduce an optimization gap between pre-training and downstream tasks that good reconstruction quality may not always lead to the high descriptive capability for the model. Considering the above issues, in this paper, we propose a simple Self-distillated masked AutoEncoder network, namely SdAE. SdAE consists of a student branch using an encoder-decoder structure to reconstruct the missing information, and a teacher branch producing latent representation of masked tokens. We also analyze how to build good views for the teacher branch to produce latent representation from the perspective of information bottleneck. After that, we propose a multi-fold masking strategy to provide multiple masked views with balanced information for boosting the performance, which can also reduce the computational complexity. Our approach generalizes well: with only 300 epochs pre-training, a vanilla ViT-Base model achieves an 84.1% fine-tuning accuracy on ImageNet-1k classification, 48.6 mIOU on ADE20K segmentation, and 48.9 mAP on COCO detection, which surpasses other methods by a considerable margin. Code is available at https://github.com/AbrahamYabo/SdAE.
Graph convolutional network based methods that model the body-joints' relations, have recently shown great promise in 3D skeleton-based human motion prediction. However, these methods have two critical issues: first, deep graph convolutions filter features within only limited graph spectrums, losing sufficient information in the full band; second, using a single graph to model the whole body underestimates the diverse patterns on various body-parts. To address the first issue, we propose adaptive graph scattering, which leverages multiple trainable band-pass graph filters to decompose pose features into richer graph spectrum bands. To address the second issue, body-parts are modeled separately to learn diverse dynamics, which enables finer feature extraction along the spatial dimensions. Integrating the above two designs, we propose a novel skeleton-parted graph scattering network (SPGSN). The cores of the model are cascaded multi-part graph scattering blocks (MPGSBs), building adaptive graph scattering on diverse body-parts, as well as fusing the decomposed features based on the inferred spectrum importance and body-part interactions. Extensive experiments have shown that SPGSN outperforms state-of-the-art methods by remarkable margins of 13.8%, 9.3% and 2.7% in terms of 3D mean per joint position error (MPJPE) on Human3.6M, CMU Mocap and 3DPW datasets, respectively.
Fine-grained object retrieval aims to learn discriminative representation to retrieve visually similar objects. However, existing top-performing works usually impose pairwise similarities on the semantic embedding spaces to continually fine-tune the entire model in limited-data regimes, thus resulting in easily converging to suboptimal solutions. In this paper, we develop Fine-grained Retrieval Prompt Tuning (FRPT), which steers a frozen pre-trained model to perform the fine-grained retrieval task from the perspectives of sample prompt and feature adaptation. Specifically, FRPT only needs to learn fewer parameters in the prompt and adaptation instead of fine-tuning the entire model, thus solving the convergence to suboptimal solutions caused by fine-tuning the entire model. Technically, as sample prompts, a structure perturbation prompt (SPP) is introduced to zoom and even exaggerate some pixels contributing to category prediction via a content-aware inhomogeneous sampling operation. In this way, SPP can make the fine-grained retrieval task aided by the perturbation prompts close to the solved task during the original pre-training. Besides, a category-specific awareness head is proposed and regarded as feature adaptation, which removes the species discrepancies in the features extracted by the pre-trained model using instance normalization, and thus makes the optimized features only include the discrepancies among subcategories. Extensive experiments demonstrate that our FRPT with fewer learnable parameters achieves the state-of-the-art performance on three widely-used fine-grained datasets.
In this paper, we present a novel protocol of annotation and evaluation for visual recognition. Different from traditional settings, the protocol does not require the labeler/algorithm to annotate/recognize all targets (objects, parts, etc.) at once, but instead raises a number of recognition instructions and the algorithm recognizes targets by request. This mechanism brings two beneficial properties to reduce the burden of annotation, namely, (i) variable granularity: different scenarios can have different levels of annotation, in particular, object parts can be labeled only in large and clear instances, (ii) being open-domain: new concepts can be added to the database in minimal costs. To deal with the proposed setting, we maintain a knowledge base and design a query-based visual recognition framework that constructs queries on-the-fly based on the requests. We evaluate the recognition system on two mixed-annotated datasets, CPP and ADE20K, and demonstrate its promising ability of learning from partially labeled data as well as adapting to new concepts with only text labels.