We present a general methodology that learns to classify images without labels by leveraging pretrained feature extractors. Our approach involves self-distillation training of clustering heads, based on the fact that nearest neighbors in the pretrained feature space are likely to share the same label. We propose a novel objective to learn associations between images by introducing a variant of pointwise mutual information together with instance weighting. We demonstrate that the proposed objective is able to attenuate the effect of false positive pairs while efficiently exploiting the structure in the pretrained feature space. As a result, we improve the clustering accuracy over $k$-means on $17$ different pretrained models by $6.1$\% and $12.2$\% on ImageNet and CIFAR100, respectively. Finally, using self-supervised pretrained vision transformers we push the clustering accuracy on ImageNet to $61.6$\%. The code will be open-sourced.
Camera motion introduces spatially varying blur due to the depth changes in the 3D world. This work investigates scene configurations where such blur is produced under parallax camera motion. We present a simple, yet accurate, Image Compositing Blur (ICB) model for depth-dependent spatially varying blur. The (forward) model produces realistic motion blur from a single image, depth map, and camera trajectory. Furthermore, we utilize the ICB model, combined with a coordinate-based MLP, to learn a sharp neural representation from the blurred input. Experimental results are reported for synthetic and real examples. The results verify that the ICB forward model is computationally efficient and produces realistic blur, despite the lack of occlusion information. Additionally, our method for restoring a sharp representation proves to be a competitive approach for the deblurring task.
Event cameras sense the intensity changes asynchronously and produce event streams with high dynamic range and low latency. This has inspired research endeavors utilizing events to guide the challenging video superresolution (VSR) task. In this paper, we make the first attempt to address a novel problem of achieving VSR at random scales by taking advantages of the high temporal resolution property of events. This is hampered by the difficulties of representing the spatial-temporal information of events when guiding VSR. To this end, we propose a novel framework that incorporates the spatial-temporal interpolation of events to VSR in a unified framework. Our key idea is to learn implicit neural representations from queried spatial-temporal coordinates and features from both RGB frames and events. Our method contains three parts. Specifically, the Spatial-Temporal Fusion (STF) module first learns the 3D features from events and RGB frames. Then, the Temporal Filter (TF) module unlocks more explicit motion information from the events near the queried timestamp and generates the 2D features. Lastly, the SpatialTemporal Implicit Representation (STIR) module recovers the SR frame in arbitrary resolutions from the outputs of these two modules. In addition, we collect a real-world dataset with spatially aligned events and RGB frames. Extensive experiments show that our method significantly surpasses the prior-arts and achieves VSR with random scales, e.g., 6.5. Code and dataset are available at https: //vlis2022.github.io/cvpr23/egvsr.
Our paper seeks to transfer the hairstyle of a reference image to an input photo for virtual hair try-on. We target a variety of challenges scenarios, such as transforming a long hairstyle with bangs to a pixie cut, which requires removing the existing hair and inferring how the forehead would look, or transferring partially visible hair from a hat-wearing person in a different pose. Past solutions leverage StyleGAN for hallucinating any missing parts and producing a seamless face-hair composite through so-called GAN inversion or projection. However, there remains a challenge in controlling the hallucinations to accurately transfer hairstyle and preserve the face shape and identity of the input. To overcome this, we propose a multi-view optimization framework that uses "two different views" of reference composites to semantically guide occluded or ambiguous regions. Our optimization shares information between two poses, which allows us to produce high fidelity and realistic results from incomplete references. Our framework produces high-quality results and outperforms prior work in a user study that consists of significantly more challenging hair transfer scenarios than previously studied. Project page: https://stylegan-salon.github.io/.
Structured, or tabular, data is the most common format in data science. While deep learning models have proven formidable in learning from unstructured data such as images or speech, they are less accurate than simpler approaches when learning from tabular data. In contrast, modern tree-based Machine Learning (ML) models shine in extracting relevant information from structured data. An essential requirement in data science is to reduce model inference latency in cases where, for example, models are used in a closed loop with simulation to accelerate scientific discovery. However, the hardware acceleration community has mostly focused on deep neural networks and largely ignored other forms of machine learning. Previous work has described the use of an analog content addressable memory (CAM) component for efficiently mapping random forests. In this work, we focus on an overall analog-digital architecture implementing a novel increased precision analog CAM and a programmable network on chip allowing the inference of state-of-the-art tree-based ML models, such as XGBoost and CatBoost. Results evaluated in a single chip at 16nm technology show 119x lower latency at 9740x higher throughput compared with a state-of-the-art GPU, with a 19W peak power consumption.
Most statistical learning algorithms rely on an over-simplified assumption, that is, the train and test data are independent and identically distributed. In real-world scenarios, however, it is common for models to encounter data from new and different domains to which they were not exposed to during training. This is often the case in medical imaging applications due to differences in acquisition devices, imaging protocols, and patient characteristics. To address this problem, domain generalization (DG) is a promising direction as it enables models to handle data from previously unseen domains by learning domain-invariant features robust to variations across different domains. To this end, we introduce a novel DG method called Adversarial Intensity Attack (AdverIN), which leverages adversarial training to generate training data with an infinite number of styles and increase data diversity while preserving essential content information. We conduct extensive evaluation experiments on various multi-domain segmentation datasets, including 2D retinal fundus optic disc/cup and 3D prostate MRI. Our results demonstrate that AdverIN significantly improves the generalization ability of the segmentation models, achieving significant improvement on these challenging datasets. Code is available upon publication.
Existing machine learning models demonstrate excellent performance in image object recognition after training on a large-scale dataset under full supervision. However, these models only learn to map an image to a predefined class index, without revealing the actual semantic meaning of the object in the image. In contrast, vision-language models like CLIP are able to assign semantic class names to unseen objects in a `zero-shot' manner, although they still rely on a predefined set of candidate names at test time. In this paper, we reconsider the recognition problem and task a vision-language model to assign class names to images given only a large and essentially unconstrained vocabulary of categories as prior information. We use non-parametric methods to establish relationships between images which allow the model to automatically narrow down the set of possible candidate names. Specifically, we propose iteratively clustering the data and voting on class names within them, showing that this enables a roughly 50\% improvement over the baseline on ImageNet. Furthermore, we tackle this problem both in unsupervised and partially supervised settings, as well as with a coarse-grained and fine-grained search space as the unconstrained dictionary.
Face swapping aims to generate swapped images that fuse the identity of source faces and the attributes of target faces. Most existing works address this challenging task through 3D modelling or generation using generative adversarial networks (GANs), but 3D modelling suffers from limited reconstruction accuracy and GANs often struggle in preserving subtle yet important identity details of source faces (e.g., skin colors, face features) and structural attributes of target faces (e.g., face shapes, facial expressions). This paper presents Face Transformer, a novel face swapping network that can accurately preserve source identities and target attributes simultaneously in the swapped face images. We introduce a transformer network for the face swapping task, which learns high-quality semantic-aware correspondence between source and target faces and maps identity features of source faces to the corresponding region in target faces. The high-quality semantic-aware correspondence enables smooth and accurate transfer of source identity information with minimal modification of target shapes and expressions. In addition, our Face Transformer incorporates a multi-scale transformation mechanism for preserving the rich fine facial details. Extensive experiments show that our Face Transformer achieves superior face swapping performance qualitatively and quantitatively.
Federated learning (FL) enables distributed clients to collaboratively train a machine learning model without sharing raw data with each other. However, it suffers the leakage of private information from uploading models. In addition, as the model size grows, the training latency increases due to limited transmission bandwidth and the model performance degrades while using differential privacy (DP) protection. In this paper, we propose a gradient sparsification empowered FL framework over wireless channels, in order to improve training efficiency without sacrificing convergence performance. Specifically, we first design a random sparsification algorithm to retain a fraction of the gradient elements in each client's local training, thereby mitigating the performance degradation induced by DP and and reducing the number of transmission parameters over wireless channels. Then, we analyze the convergence bound of the proposed algorithm, by modeling a non-convex FL problem. Next, we formulate a time-sequential stochastic optimization problem for minimizing the developed convergence bound, under the constraints of transmit power, the average transmitting delay, as well as the client's DP requirement. Utilizing the Lyapunov drift-plus-penalty framework, we develop an analytical solution to the optimization problem. Extensive experiments have been implemented on three real life datasets to demonstrate the effectiveness of our proposed algorithm. We show that our proposed algorithms can fully exploit the interworking between communication and computation to outperform the baselines, i.e., random scheduling, round robin and delay-minimization algorithms.
Although fast adversarial training provides an efficient approach for building robust networks, it may suffer from a serious problem known as catastrophic overfitting (CO), where the multi-step robust accuracy suddenly collapses to zero. In this paper, we for the first time decouple the FGSM examples into data-information and self-information, which reveals an interesting phenomenon called "self-fitting". Self-fitting, i.e., DNNs learn the self-information embedded in single-step perturbations, naturally leads to the occurrence of CO. When self-fitting occurs, the network experiences an obvious "channel differentiation" phenomenon that some convolution channels accounting for recognizing self-information become dominant, while others for data-information are suppressed. In this way, the network learns to only recognize images with sufficient self-information and loses generalization ability to other types of data. Based on self-fitting, we provide new insight into the existing methods to mitigate CO and extend CO to multi-step adversarial training. Our findings reveal a self-learning mechanism in adversarial training and open up new perspectives for suppressing different kinds of information to mitigate CO.