Food image analysis is the groundwork for image-based dietary assessment, which is the process of monitoring what kinds of food and how much energy is consumed using captured food or eating scene images. Existing deep learning-based methods learn the visual representation for downstream tasks based on human annotation of each food image. However, most food images in real life are obtained without labels, and data annotation requires plenty of time and human effort, which is not feasible for real-world applications. To make use of the vast amount of unlabeled images, many existing works focus on unsupervised or self-supervised learning of visual representations directly from unlabeled data. However, none of these existing works focus on food images, which is more challenging than general objects due to its high inter-class similarity and intra-class variance. In this paper, we focus on the implementation and analysis of existing representative self-supervised learning methods on food images. Specifically, we first compare the performance of six selected self-supervised learning models on the Food-101 dataset. Then we analyze the pros and cons of each selected model when training on food data to identify the key factors that can help improve the performance. Finally, we propose several ideas for future work on self-supervised visual representation learning for food images.
Shadow removal in a single image has received increasing attention in recent years. However, removing shadows over dynamic scenes remains largely under-explored. In this paper, we propose the first data-driven video shadow removal model, termed PSTNet, by exploiting three essential characteristics of video shadows, i.e., physical property, spatio relation, and temporal coherence. Specifically, a dedicated physical branch was established to conduct local illumination estimation, which is more applicable for scenes with complex lighting and textures, and then enhance the physical features via a mask-guided attention strategy. Then, we develop a progressive aggregation module to enhance the spatio and temporal characteristics of features maps, and effectively integrate the three kinds of features. Furthermore, to tackle the lack of datasets of paired shadow videos, we synthesize a dataset (SVSRD-85) with aid of the popular game GTAV by controlling the switch of the shadow renderer. Experiments against 9 state-of-the-art models, including image shadow removers and image/video restoration methods, show that our method improves the best SOTA in terms of RMSE error for the shadow area by 14.7. In addition, we develop a lightweight model adaptation strategy to make our synthetic-driven model effective in real world scenes. The visual comparison on the public SBU-TimeLapse dataset verifies the generalization ability of our model in real scenes.
Learning generalizable policies that can adapt to unseen environments remains challenging in visual Reinforcement Learning (RL). Existing approaches try to acquire a robust representation via diversifying the appearances of in-domain observations for better generalization. Limited by the specific observations of the environment, these methods ignore the possibility of exploring diverse real-world image datasets. In this paper, we investigate how a visual RL agent would benefit from the off-the-shelf visual representations. Surprisingly, we find that the early layers in an ImageNet pre-trained ResNet model could provide rather generalizable representations for visual RL. Hence, we propose Pre-trained Image Encoder for Generalizable visual reinforcement learning (PIE-G), a simple yet effective framework that can generalize to the unseen visual scenarios in a zero-shot manner. Extensive experiments are conducted on DMControl Generalization Benchmark, DMControl Manipulation Tasks, Drawer World, and CARLA to verify the effectiveness of PIE-G. Empirical evidence suggests PIE-G improves sample efficiency and significantly outperforms previous state-of-the-art methods in terms of generalization performance. In particular, PIE-G boasts a 55% generalization performance gain on average in the challenging video background setting. Project Page: https://sites.google.com/view/pie-g/home.
Few-shot class-incremental learning (FSCIL) has recently attracted extensive attention in various areas. Existing FSCIL methods highly depend on the robustness of the feature backbone pre-trained on base classes. In recent years, different Transformer variants have obtained significant processes in the feature representation learning of massive fields. Nevertheless, the progress of the Transformer in FSCIL scenarios has not achieved the potential promised in other fields so far. In this paper, we develop a semantic-visual guided Transformer (SV-T) to enhance the feature extracting capacity of the pre-trained feature backbone on incremental classes. Specifically, we first utilize the visual (image) labels provided by the base classes to supervise the optimization of the Transformer. And then, a text encoder is introduced to automatically generate the corresponding semantic (text) labels for each image from the base classes. Finally, the constructed semantic labels are further applied to the Transformer for guiding its hyperparameters updating. Our SV-T can take full advantage of more supervision information from base classes and further enhance the training robustness of the feature backbone. More importantly, our SV-T is an independent method, which can directly apply to the existing FSCIL architectures for acquiring embeddings of various incremental classes. Extensive experiments on three benchmarks, two FSCIL architectures, and two Transformer variants show that our proposed SV-T obtains a significant improvement in comparison to the existing state-of-the-art FSCIL methods.
Performance of trimap-free image matting methods is limited when trying to decouple the deterministic and undetermined regions, especially in the scenes where foregrounds are semantically ambiguous, chromaless, or high transmittance. In this paper, we propose a novel framework named Privileged Prior Information Distillation for Image Matting (PPID-IM) that can effectively transfer privileged prior environment-aware information to improve the performance of students in solving hard foregrounds. The prior information of trimap regulates only the teacher model during the training stage, while not being fed into the student network during actual inference. In order to achieve effective privileged cross-modality (i.e. trimap and RGB) information distillation, we introduce a Cross-Level Semantic Distillation (CLSD) module that reinforces the trimap-free students with more knowledgeable semantic representations and environment-aware information. We also propose an Attention-Guided Local Distillation module that efficiently transfers privileged local attributes from the trimap-based teacher to trimap-free students for the guidance of local-region optimization. Extensive experiments demonstrate the effectiveness and superiority of our PPID framework on the task of image matting. In addition, our trimap-free IndexNet-PPID surpasses the other competing state-of-the-art methods by a large margin, especially in scenarios with chromaless, weak texture, or irregular objects.
Neural Architecture Search (NAS) for automatically finding the optimal network architecture has shown some success with competitive performances in various computer vision tasks. However, NAS in general requires a tremendous amount of computations. Thus reducing computational cost has emerged as an important issue. Most of the attempts so far has been based on manual approaches, and often the architectures developed from such efforts dwell in the balance of the network optimality and the search cost. Additionally, recent NAS methods for image restoration generally do not consider dynamic operations that may transform dimensions of feature maps because of the dimensionality mismatch in tensor calculations. This can greatly limit NAS in its search for optimal network structure. To address these issues, we re-frame the optimal search problem by focusing at component block level. From previous work, it's been shown that an effective denoising block can be connected in series to further improve the network performance. By focusing at block level, the search space of reinforcement learning becomes significantly smaller and evaluation process can be conducted more rapidly. In addition, we integrate an innovative dimension matching modules for dealing with spatial and channel-wise mismatch that may occur in the optimal design search. This allows much flexibility in optimal network search within the cell block. With these modules, then we employ reinforcement learning in search of an optimal image denoising network at a module level. Computational efficiency of our proposed Denoising Prior Neural Architecture Search (DPNAS) was demonstrated by having it complete an optimal architecture search for an image restoration task by just one day with a single GPU.
We qualitatively examine the accuracy and fideltiy between two diffusion-based image generation systems, namely DALL-E 2 and Luna, which have massive differences in training datasets, algorithmic approaches, prompt resolvement, and output upscaling. In our research we conclude that DALL-E 2 significantly edges Luna in both alignment and fidelity comparisons
While current multi-frame restoration methods combine information from multiple input images using 2D alignment techniques, recent advances in novel view synthesis are paving the way for a new paradigm relying on volumetric scene representations. In this work, we introduce the first 3D-based multi-frame denoising method that significantly outperforms its 2D-based counterparts with lower computational requirements. Our method extends the multiplane image (MPI) framework for novel view synthesis by introducing a learnable encoder-renderer pair manipulating multiplane representations in feature space. The encoder fuses information across views and operates in a depth-wise manner while the renderer fuses information across depths and operates in a view-wise manner. The two modules are trained end-to-end and learn to separate depths in an unsupervised way, giving rise to Multiplane Feature (MPF) representations. Experiments on the Spaces and Real Forward-Facing datasets as well as on raw burst data validate our approach for view synthesis, multi-frame denoising, and view synthesis under noisy conditions.
Self-supervised pretraining has been observed to improve performance in supervised learning tasks in medical imaging. This study investigates the utility of self-supervised pretraining prior to conducting supervised fine-tuning for the downstream task of lung sliding classification in M-mode lung ultrasound images. We propose a novel pairwise relationship that couples M-mode images constructed from the same B-mode image and investigate the utility of data augmentation procedure specific to M-mode lung ultrasound. The results indicate that self-supervised pretraining yields better performance than full supervision, most notably for feature extractors not initialized with ImageNet-pretrained weights. Moreover, we observe that including a vast volume of unlabelled data results in improved performance on external validation datasets, underscoring the value of self-supervision for improving generalizability in automatic ultrasound interpretation. To the authors' best knowledge, this study is the first to characterize the influence of self-supervised pretraining for M-mode ultrasound.
We propose a novel type of map for visual navigation, a renderable neural radiance map (RNR-Map), which is designed to contain the overall visual information of a 3D environment. The RNR-Map has a grid form and consists of latent codes at each pixel. These latent codes are embedded from image observations, and can be converted to the neural radiance field which enables image rendering given a camera pose. The recorded latent codes implicitly contain visual information about the environment, which makes the RNR-Map visually descriptive. This visual information in RNR-Map can be a useful guideline for visual localization and navigation. We develop localization and navigation frameworks that can effectively utilize the RNR-Map. We evaluate the proposed frameworks on camera tracking, visual localization, and image-goal navigation. Experimental results show that the RNR-Map-based localization framework can find the target location based on a single query image with fast speed and competitive accuracy compared to other baselines. Also, this localization framework is robust to environmental changes, and even finds the most visually similar places when a query image from a different environment is given. The proposed navigation framework outperforms the existing image-goal navigation methods in difficult scenarios, under odometry and actuation noises. The navigation framework shows 65.7% success rate in curved scenarios of the NRNS dataset, which is an improvement of 18.6% over the current state-of-the-art.