GRIT: General Robust Image Task Benchmark

Computer vision models excel at making predictions when the test distribution closely resembles the training distribution. Such models have yet to match the ability of biological vision to learn from multiple sources and generalize to new data sources and tasks. To facilitate the development and evaluation of more general vision systems, we introduce the General Robust Image Task (GRIT) benchmark. GRIT evaluates the performance, robustness, and calibration of a vision system across a variety of image prediction tasks, concepts, and data sources. The seven tasks in GRIT are selected to cover a range of visual skills: object categorization, object localization, referring expression grounding, visual question answering, segmentation, human keypoint detection, and surface normal estimation. GRIT is carefully designed to enable the evaluation of robustness under image perturbations, image source distribution shift, and concept distribution shift. By providing a unified platform for thorough assessment of skills and concepts learned by a vision model, we hope GRIT catalyzes the development of performant and robust general purpose vision systems.

Shape-Guided Diffusion with Inside-Outside Attention

Shape can specify key object constraints, yet existing text-to-image diffusion models ignore this cue and synthesize objects that are incorrectly scaled, cut off, or replaced with background content. We propose a training-free method, Shape-Guided Diffusion, which uses a novel Inside-Outside Attention mechanism to constrain the cross-attention (and self-attention) maps such that prompt tokens (and pixels) referring to the inside of the shape cannot attend outside the shape, and vice versa. To demonstrate the efficacy of our method, we propose a new image editing task where the model must replace an object specified by its mask and a text prompt. We curate a new ShapePrompts benchmark based on MS-COCO and achieve SOTA results in shape faithfulness, text alignment, and realism according to both quantitative metrics and human preferences. Our data and code will be made available at https://shape-guided-diffusion.github.io.

RLIP: Relational Language-Image Pre-training for Human-Object Interaction Detection

The task of Human-Object Interaction (HOI) detection targets fine-grained visual parsing of humans interacting with their environment, enabling a broad range of applications. Prior work has demonstrated the benefits of effective architecture design and integration of relevant cues for more accurate HOI detection. However, the design of an appropriate pre-training strategy for this task remains underexplored by existing approaches. To address this gap, we propose Relational Language-Image Pre-training (RLIP), a strategy for contrastive pre-training that leverages both entity and relation descriptions. To make effective use of such pre-training, we make three technical contributions: (1) a new Parallel entity detection and Sequential relation inference (ParSe) architecture that enables the use of both entity and relation descriptions during holistically optimized pre-training; (2) a synthetic data generation framework, Label Sequence Extension, that expands the scale of language data available within each minibatch; (3) mechanisms to account for ambiguity, Relation Quality Labels and Relation Pseudo-Labels, to mitigate the influence of ambiguous/noisy samples in the pre-training data. Through extensive experiments, we demonstrate the benefits of these contributions, collectively termed RLIP-ParSe, for improved zero-shot, few-shot and fine-tuning HOI detection performance as well as increased robustness to learning from noisy annotations. Code will be available at \url{https://github.com/JacobYuan7/RLIP}.

Decay2Distill: Leveraging spatial perturbation and regularization for self-supervised image denoising

Unpaired image denoising has achieved promising development over the last few years. Regardless of the performance, methods tend to heavily rely on underlying noise properties or any assumption which is not always practical. Alternatively, if we can ground the problem from a structural perspective rather than noise statistics, we can achieve a more robust solution. with such motivation, we propose a self-supervised denoising scheme that is unpaired and relies on spatial degradation followed by a regularized refinement. Our method shows considerable improvement over previous methods and exhibited consistent performance over different data domains.

CapOnImage: Context-driven Dense-Captioning on Image

Existing image captioning systems are dedicated to generating narrative captions for images, which are spatially detached from the image in presentation. However, texts can also be used as decorations on the image to highlight the key points and increase the attractiveness of images. In this work, we introduce a new task called captioning on image (CapOnImage), which aims to generate dense captions at different locations of the image based on contextual information. To fully exploit the surrounding visual context to generate the most suitable caption for each location, we propose a multi-modal pre-training model with multi-level pre-training tasks that progressively learn the correspondence between texts and image locations from easy to difficult. Since the model may generate redundant captions for nearby locations, we further enhance the location embedding with neighbor locations as context. For this new task, we also introduce a large-scale benchmark called CapOnImage2M, which contains 2.1 million product images, each with an average of 4.8 spatially localized captions. Compared with other image captioning model variants, our model achieves the best results in both captioning accuracy and diversity aspects. We will make code and datasets public to facilitate future research.

Evaluating the Practicality of Learned Image Compression

Learned image compression has achieved extraordinary rate-distortion performance in PSNR and MS-SSIM compared to traditional methods. However, it suffers from intensive computation, which is intolerable for real-world applications and leads to its limited industrial application for now. In this paper, we introduce neural architecture search (NAS) to designing more efficient networks with lower latency, and leverage quantization to accelerate the inference process. Meanwhile, efforts in engineering like multi-threading and SIMD have been made to improve efficiency. Optimized using a hybrid loss of PSNR and MS-SSIM for better visual quality, we obtain much higher MS-SSIM than JPEG, JPEG XL and AVIF over all bit rates, and PSNR between that of JPEG XL and AVIF. Our software implementation of LIC achieves comparable or even faster inference speed compared to jpeg-turbo while being multiple times faster than JPEG XL and AVIF. Besides, our implementation of LIC reaches stunning throughput of 145 fps for encoding and 208 fps for decoding on a Tesla T4 GPU for 1080p images. On CPU, the latency of our implementation is comparable with JPEG XL.

Learning Detail-Structure Alternative Optimization for Blind Super-Resolution

Existing convolutional neural networks (CNN) based image super-resolution (SR) methods have achieved impressive performance on bicubic kernel, which is not valid to handle unknown degradations in real-world applications. Recent blind SR methods suggest to reconstruct SR images relying on blur kernel estimation. However, their results still remain visible artifacts and detail distortion due to the estimation errors. To alleviate these problems, in this paper, we propose an effective and kernel-free network, namely DSSR, which enables recurrent detail-structure alternative optimization without blur kernel prior incorporation for blind SR. Specifically, in our DSSR, a detail-structure modulation module (DSMM) is built to exploit the interaction and collaboration of image details and structures. The DSMM consists of two components: a detail restoration unit (DRU) and a structure modulation unit (SMU). The former aims at regressing the intermediate HR detail reconstruction from LR structural contexts, and the latter performs structural contexts modulation conditioned on the learned detail maps at both HR and LR spaces. Besides, we use the output of DSMM as the hidden state and design our DSSR architecture from a recurrent convolutional neural network (RCNN) view. In this way, the network can alternatively optimize the image details and structural contexts, achieving co-optimization across time. Moreover, equipped with the recurrent connection, our DSSR allows low- and high-level feature representations complementary by observing previous HR details and contexts at every unrolling time. Extensive experiments on synthetic datasets and real-world images demonstrate that our method achieves the state-of-the-art against existing methods. The source code can be found at https://github.com/Arcananana/DSSR.

Object Delineation in Satellite Images

Machine learning is being widely applied to analyze satellite data with problems such as classification and feature detection. Unlike traditional image processing algorithms, geospatial applications need to convert the detected objects from a raster form to a geospatial vector form to further analyze it. This gem delivers a simple and light-weight algorithm for delineating the pixels that are marked by ML algorithms to extract geospatial objects from satellite images. The proposed algorithm is exact and users can further apply simplification and approximation based on the application needs.

Pixel-by-pixel Mean Opinion Score (pMOS) for No-Reference Image Quality Assessment

Deep-learning based techniques have contributed to the remarkable progress in the field of automatic image quality assessment (IQA). Existing IQA methods are designed to measure the quality of an image in terms of Mean Opinion Score (MOS) at the image-level (i.e. the whole image) or at the patch-level (dividing the image into multiple units and measuring quality of each patch). Some applications may require assessing the quality at the pixel-level (i.e. MOS value for each pixel), however, this is not possible in case of existing techniques as the spatial information is lost owing to their network structures. This paper proposes an IQA algorithm that can measure the MOS at the pixel-level, in addition to the image-level MOS. The proposed algorithm consists of three core parts, namely: i) Local IQA; ii) Region of Interest (ROI) prediction; iii) High-level feature embedding. The Local IQA part outputs the MOS at the pixel-level, or pixel-by-pixel MOS - we term it 'pMOS'. The ROI prediction part outputs weights that characterize the relative importance of region when calculating the image-level IQA. The high-level feature embedding part extracts high-level image features which are then embedded into the Local IQA part. In other words, the proposed algorithm yields three outputs: the pMOS which represents MOS for each pixel, the weights from the ROI indicating the relative importance of region, and finally the image-level MOS that is obtained by the weighted sum of pMOS and ROI values. The image-level MOS thus obtained by utilizing pMOS and ROI weights shows superior performance compared to the existing popular IQA techniques. In addition, visualization results indicate that predicted pMOS and ROI outputs are reasonably aligned with the general principles of the human visual system (HVS).

Nervus: A Comprehensive Deep Learning Classification, Regression, and Prognostication Tool for both Medical Image and Clinical Data Analysis

The goal of our research is to create a comprehensive and flexible library that is easy to use for medical imaging research, and capable of handling grayscale images, multiple inputs (both images and tabular data), and multi-label tasks. We have named it Nervus. Based on the PyTorch library, which is suitable for AI for research purposes, we created a four-part model to handle comprehensive inputs and outputs. Nervus consists of four parts. First is the dataloader, then the feature extractor, the feature mixer, and finally the classifier. The dataloader preprocesses the input data, the feature extractor extracts the features between the training data and ground truth labels, feature mixer mixes the features of the extractors, and the classifier classifies the input data from feature mixer based on the task. We have created Nervus, which is a comprehensive and flexible model library that is easy to use for medical imaging research which can handle grayscale images, multi-inputs and multi-label tasks. This will be helpful for researchers in the field of radiology.