Caner Hazirbas

Zhiheng Li, Ivan Evtimov, Albert Gordo et al. · amazon-science

Machine learning models have been found to learn shortcuts -- unintended decision rules that are unable to generalize -- undermining models' reliability. Previous works address this problem under the tenuous assumption that only a single shortcut exists in the training data. Real-world images are rife with multiple visual cues from background to texture. Key to advancing the reliability of vision systems is understanding whether existing methods can overcome multiple shortcuts or struggle in a Whac-A-Mole game, i.e., where mitigating one shortcut amplifies reliance on others. To address this shortcoming, we propose two benchmarks: 1) UrbanCars, a dataset with precisely controlled spurious cues, and 2) ImageNet-W, an evaluation set based on ImageNet for watermark, a shortcut we discovered affects nearly every modern vision model. Along with texture and background, ImageNet-W allows us to study multiple shortcuts emerging from training on natural images. We find computer vision models, including large foundation models -- regardless of training set, architecture, and supervision -- struggle when multiple shortcuts are present. Even methods explicitly designed to combat shortcuts struggle in a Whac-A-Mole dilemma. To tackle this challenge, we propose Last Layer Ensemble, a simple-yet-effective method to mitigate multiple shortcuts without Whac-A-Mole behavior. Our results surface multi-shortcut mitigation as an overlooked challenge critical to advancing the reliability of vision systems. The datasets and code are released: https://github.com/facebookresearch/Whac-A-Mole.

Badr Youbi Idrissi, Diane Bouchacourt, Randall Balestriero et al. · meta-ai

Deep learning vision systems are widely deployed across applications where reliability is critical. However, even today's best models can fail to recognize an object when its pose, lighting, or background varies. While existing benchmarks surface examples challenging for models, they do not explain why such mistakes arise. To address this need, we introduce ImageNet-X, a set of sixteen human annotations of factors such as pose, background, or lighting the entire ImageNet-1k validation set as well as a random subset of 12k training images. Equipped with ImageNet-X, we investigate 2,200 current recognition models and study the types of mistakes as a function of model's (1) architecture, e.g. transformer vs. convolutional, (2) learning paradigm, e.g. supervised vs. self-supervised, and (3) training procedures, e.g., data augmentation. Regardless of these choices, we find models have consistent failure modes across ImageNet-X categories. We also find that while data augmentation can improve robustness to certain factors, they induce spill-over effects to other factors. For example, strong random cropping hurts robustness on smaller objects. Together, these insights suggest to advance the robustness of modern vision models, future research should focus on collecting additional data and understanding data augmentation schemes. Along with these insights, we release a toolkit based on ImageNet-X to spur further study into the mistakes image recognition systems make.

Jiachen Sun, Mark Ibrahim, Melissa Hall et al.

Textual prompt tuning has demonstrated significant performance improvements in adapting natural language processing models to a variety of downstream tasks by treating hand-engineered prompts as trainable parameters. Inspired by the success of textual prompting, several studies have investigated the efficacy of visual prompt tuning. In this work, we present Visual Prompt Adaptation (VPA), the first framework that generalizes visual prompting with test-time adaptation. VPA introduces a small number of learnable tokens, enabling fully test-time and storage-efficient adaptation without necessitating source-domain information. We examine our VPA design under diverse adaptation settings, encompassing single-image, batched-image, and pseudo-label adaptation. We evaluate VPA on multiple tasks, including out-of-distribution (OOD) generalization, corruption robustness, and domain adaptation. Experimental results reveal that VPA effectively enhances OOD generalization by 3.3% across various models, surpassing previous test-time approaches. Furthermore, we show that VPA improves corruption robustness by 6.5% compared to strong baselines. Finally, we demonstrate that VPA also boosts domain adaptation performance by relatively 5.2%. Our VPA also exhibits marked effectiveness in improving the robustness of zero-shot recognition for vision-language models.

Haider Al-Tahan, Quentin Garrido, Randall Balestriero et al.

Significant research efforts have been made to scale and improve vision-language model (VLM) training approaches. Yet, with an ever-growing number of benchmarks, researchers are tasked with the heavy burden of implementing each protocol, bearing a non-trivial computational cost, and making sense of how all these benchmarks translate into meaningful axes of progress. To facilitate a systematic evaluation of VLM progress, we introduce UniBench: a unified implementation of 50+ VLM benchmarks spanning a comprehensive range of carefully categorized capabilities from object recognition to spatial awareness, counting, and much more. We showcase the utility of UniBench for measuring progress by evaluating nearly 60 publicly available vision-language models, trained on scales of up to 12.8B samples. We find that while scaling training data or model size can boost many vision-language model capabilities, scaling offers little benefit for reasoning or relations. Surprisingly, we also discover today's best VLMs struggle on simple digit recognition and counting tasks, e.g. MNIST, which much simpler networks can solve. Where scale falls short, we find that more precise interventions, such as data quality or tailored-learning objectives offer more promise. For practitioners, we also offer guidance on selecting a suitable VLM for a given application. Finally, we release an easy-to-run UniBench code-base with the full set of 50+ benchmarks and comparisons across 59 models as well as a distilled, representative set of benchmarks that runs in 5 minutes on a single GPU.

Laura Gustafson, Megan Richards, Melissa Hall et al.

Despite impressive advances in object-recognition, deep learning systems' performance degrades significantly across geographies and lower income levels raising pressing concerns of inequity. Addressing such performance gaps remains a challenge, as little is understood about why performance degrades across incomes or geographies. We take a step in this direction by annotating images from Dollar Street, a popular benchmark of geographically and economically diverse images, labeling each image with factors such as color, shape, and background. These annotations unlock a new granular view into how objects differ across incomes and regions. We then use these object differences to pinpoint model vulnerabilities across incomes and regions. We study a range of modern vision models, finding that performance disparities are most associated with differences in texture, occlusion, and images with darker lighting. We illustrate how insights from our factor labels can surface mitigations to improve models' performance disparities. As an example, we show that mitigating a model's vulnerability to texture can improve performance on the lower income level. We release all the factor annotations along with an interactive dashboard to facilitate research into more equitable vision systems.

Caner Hazirbas, Yejin Bang, Tiezheng Yu et al.

Developing robust and fair AI systems require datasets with comprehensive set of labels that can help ensure the validity and legitimacy of relevant measurements. Recent efforts, therefore, focus on collecting person-related datasets that have carefully selected labels, including sensitive characteristics, and consent forms in place to use those attributes for model testing and development. Responsible data collection involves several stages, including but not limited to determining use-case scenarios, selecting categories (annotations) such that the data are fit for the purpose of measuring algorithmic bias for subgroups and most importantly ensure that the selected categories/subcategories are robust to regional diversities and inclusive of as many subgroups as possible. Meta, in a continuation of our efforts to measure AI algorithmic bias and robustness (https://ai.facebook.com/blog/shedding-light-on-fairness-in-ai-with-a-new-data-set), is working on collecting a large consent-driven dataset with a comprehensive list of categories. This paper describes our proposed design of such categories and subcategories for Casual Conversations v2.

Bilal Porgali, Vítor Albiero, Jordan Ryda et al.

This paper introduces a new large consent-driven dataset aimed at assisting in the evaluation of algorithmic bias and robustness of computer vision and audio speech models in regards to 11 attributes that are self-provided or labeled by trained annotators. The dataset includes 26,467 videos of 5,567 unique paid participants, with an average of almost 5 videos per person, recorded in Brazil, India, Indonesia, Mexico, Vietnam, Philippines, and the USA, representing diverse demographic characteristics. The participants agreed for their data to be used in assessing fairness of AI models and provided self-reported age, gender, language/dialect, disability status, physical adornments, physical attributes and geo-location information, while trained annotators labeled apparent skin tone using the Fitzpatrick Skin Type and Monk Skin Tone scales, and voice timbre. Annotators also labeled for different recording setups and per-second activity annotations.

Maxim Maximov, Tim Meinhardt, Ismail Elezi et al.

The advent of data-driven technology solutions is accompanied by an increasing concern with data privacy. This is of particular importance for human-centered image recognition tasks, such as pedestrian detection, re-identification, and tracking. To highlight the importance of privacy issues and motivate future research, we motivate and introduce the Pedestrian Dataset De-Identification (PDI) task. PDI evaluates the degree of de-identification and downstream task training performance for a given de-identification method. As a first baseline, we propose IncogniMOT, a two-stage full-body de-identification pipeline based on image synthesis via generative adversarial networks. The first stage replaces target pedestrians with synthetic identities. To improve downstream task performance, we then apply stage two, which blends and adapts the synthetic image parts into the data. To demonstrate the effectiveness of IncogniMOT, we generate a fully de-identified version of the MOT17 pedestrian tracking dataset and analyze its application as training data for pedestrian re-identification, detection, and tracking models. Furthermore, we show how our data is able to narrow the synthetic-to-real performance gap in a privacy-conscious manner.

Caner Hazirbas, Alicia Sun, Yonathan Efroni et al.

Despite the remarkable performance of foundation vision-language models, the shared representation space for text and vision can also encode harmful label associations detrimental to fairness. While prior work has uncovered bias in vision-language models' (VLMs) classification performance across geography, work has been limited along the important axis of harmful label associations due to a lack of rich, labeled data. In this work, we investigate harmful label associations in the recently released Casual Conversations datasets containing more than 70,000 videos. We study bias in the frequency of harmful label associations across self-provided labels for age, gender, apparent skin tone, and physical adornments across several leading VLMs. We find that VLMs are $4-7$x more likely to harmfully classify individuals with darker skin tones. We also find scaling transformer encoder model size leads to higher confidence in harmful predictions. Finally, we find improvements on standard vision tasks across VLMs does not address disparities in harmful label associations.

Vikash Sehwag, Caner Hazirbas, Albert Gordo et al.

Our work focuses on addressing sample deficiency from low-density regions of data manifold in common image datasets. We leverage diffusion process based generative models to synthesize novel images from low-density regions. We observe that uniform sampling from diffusion models predominantly samples from high-density regions of the data manifold. Therefore, we modify the sampling process to guide it towards low-density regions while simultaneously maintaining the fidelity of synthetic data. We rigorously demonstrate that our process successfully generates novel high fidelity samples from low-density regions. We further examine generated samples and show that the model does not memorize low-density data and indeed learns to generate novel samples from low-density regions.

Priya Goyal, Adriana Romero Soriano, Caner Hazirbas et al.

Does everyone equally benefit from computer vision systems? Answers to this question become more and more important as computer vision systems are deployed at large scale, and can spark major concerns when they exhibit vast performance discrepancies between people from various demographic and social backgrounds. Systematic diagnosis of fairness, harms, and biases of computer vision systems is an important step towards building socially responsible systems. To initiate an effort towards standardized fairness audits, we propose three fairness indicators, which aim at quantifying harms and biases of visual systems. Our indicators use existing publicly available datasets collected for fairness evaluations, and focus on three main types of harms and bias identified in the literature, namely harmful label associations, disparity in learned representations of social and demographic traits, and biased performance on geographically diverse images from across the world.We define precise experimental protocols applicable to a wide range of computer vision models. These indicators are part of an ever-evolving suite of fairness probes and are not intended to be a substitute for a thorough analysis of the broader impact of the new computer vision technologies. Yet, we believe it is a necessary first step towards (1) facilitating the widespread adoption and mandate of the fairness assessments in computer vision research, and (2) tracking progress towards building socially responsible models. To study the practical effectiveness and broad applicability of our proposed indicators to any visual system, we apply them to off-the-shelf models built using widely adopted model training paradigms which vary in their ability to whether they can predict labels on a given image or only produce the embeddings. We also systematically study the effect of data domain and model size.

Chunxi Liu, Michael Picheny, Leda Sarı et al.

It is well known that many machine learning systems demonstrate bias towards specific groups of individuals. This problem has been studied extensively in the Facial Recognition area, but much less so in Automatic Speech Recognition (ASR). This paper presents initial Speech Recognition results on "Casual Conversations" -- a publicly released 846 hour corpus designed to help researchers evaluate their computer vision and audio models for accuracy across a diverse set of metadata, including age, gender, and skin tone. The entire corpus has been manually transcribed, allowing for detailed ASR evaluations across these metadata. Multiple ASR models are evaluated, including models trained on LibriSpeech, 14,000 hour transcribed, and over 2 million hour untranscribed social media videos. Significant differences in word error rate across gender and skin tone are observed at times for all models. We are releasing human transcripts from the Casual Conversations dataset to encourage the community to develop a variety of techniques to reduce these statistical biases.

Ousmane Amadou Dia, Theofanis Karaletsos, Caner Hazirbas et al.

The susceptibility of deep learning models to adversarial perturbations has stirred renewed attention in adversarial examples resulting in a number of attacks. However, most of these attacks fail to encompass a large spectrum of adversarial perturbations that are imperceptible to humans. In this paper, we present localized uncertainty attacks, a novel class of threat models against deterministic and stochastic classifiers. Under this threat model, we create adversarial examples by perturbing only regions in the inputs where a classifier is uncertain. To find such regions, we utilize the predictive uncertainty of the classifier when the classifier is stochastic or, we learn a surrogate model to amortize the uncertainty when it is deterministic. Unlike $\ell_p$ ball or functional attacks which perturb inputs indiscriminately, our targeted changes can be less perceptible. When considered under our threat model, these attacks still produce strong adversarial examples; with the examples retaining a greater degree of similarity with the inputs.

Caner Hazirbas, Joanna Bitton, Brian Dolhansky et al.

This paper introduces a novel dataset to help researchers evaluate their computer vision and audio models for accuracy across a diverse set of age, genders, apparent skin tones and ambient lighting conditions. Our dataset is composed of 3,011 subjects and contains over 45,000 videos, with an average of 15 videos per person. The videos were recorded in multiple U.S. states with a diverse set of adults in various age, gender and apparent skin tone groups. A key feature is that each subject agreed to participate for their likenesses to be used. Additionally, our age and gender annotations are provided by the subjects themselves. A group of trained annotators labeled the subjects' apparent skin tone using the Fitzpatrick skin type scale. Moreover, annotations for videos recorded in low ambient lighting are also provided. As an application to measure robustness of predictions across certain attributes, we provide a comprehensive study on the top five winners of the DeepFake Detection Challenge (DFDC). Experimental evaluation shows that the winning models are less performant on some specific groups of people, such as subjects with darker skin tones and thus may not generalize to all people. In addition, we also evaluate the state-of-the-art apparent age and gender classification methods. Our experiments provides a thorough analysis on these models in terms of fair treatment of people from various backgrounds.

Nikolaus Mayer, Eddy Ilg, Philipp Fischer et al.

The finding that very large networks can be trained efficiently and reliably has led to a paradigm shift in computer vision from engineered solutions to learning formulations. As a result, the research challenge shifts from devising algorithms to creating suitable and abundant training data for supervised learning. How to efficiently create such training data? The dominant data acquisition method in visual recognition is based on web data and manual annotation. Yet, for many computer vision problems, such as stereo or optical flow estimation, this approach is not feasible because humans cannot manually enter a pixel-accurate flow field. In this paper, we promote the use of synthetically generated data for the purpose of training deep networks on such tasks.We suggest multiple ways to generate such data and evaluate the influence of dataset properties on the performance and generalization properties of the resulting networks. We also demonstrate the benefit of learning schedules that use different types of data at selected stages of the training process.

Tim Meinhardt, Michael Moeller, Caner Hazirbas et al.

While variational methods have been among the most powerful tools for solving linear inverse problems in imaging, deep (convolutional) neural networks have recently taken the lead in many challenging benchmarks. A remaining drawback of deep learning approaches is their requirement for an expensive retraining whenever the specific problem, the noise level, noise type, or desired measure of fidelity changes. On the contrary, variational methods have a plug-and-play nature as they usually consist of separate data fidelity and regularization terms. In this paper we study the possibility of replacing the proximal operator of the regularization used in many convex energy minimization algorithms by a denoising neural network. The latter therefore serves as an implicit natural image prior, while the data term can still be chosen independently. Using a fixed denoising neural network in exemplary problems of image deconvolution with different blur kernels and image demosaicking, we obtain state-of-the-art reconstruction results. These indicate the high generalizability of our approach and a reduction of the need for problem-specific training. Additionally, we discuss novel results on the analysis of possible optimization algorithms to incorporate the network into, as well as the choices of algorithm parameters and their relation to the noise level the neural network is trained on.

Caner Hazirbas, Sebastian Georg Soyer, Maximilian Christian Staab et al.

Depth from focus (DFF) is one of the classical ill-posed inverse problems in computer vision. Most approaches recover the depth at each pixel based on the focal setting which exhibits maximal sharpness. Yet, it is not obvious how to reliably estimate the sharpness level, particularly in low-textured areas. In this paper, we propose `Deep Depth From Focus (DDFF)' as the first end-to-end learning approach to this problem. One of the main challenges we face is the hunger for data of deep neural networks. In order to obtain a significant amount of focal stacks with corresponding groundtruth depth, we propose to leverage a light-field camera with a co-calibrated RGB-D sensor. This allows us to digitally create focal stacks of varying sizes. Compared to existing benchmarks our dataset is 25 times larger, enabling the use of machine learning for this inverse problem. We compare our results with state-of-the-art DFF methods and we also analyze the effect of several key deep architectural components. These experiments show that our proposed method `DDFFNet' achieves state-of-the-art performance in all scenes, reducing depth error by more than 75% compared to the classical DFF methods.

Florian Walch, Caner Hazirbas, Laura Leal-Taixé et al.

In this work we propose a new CNN+LSTM architecture for camera pose regression for indoor and outdoor scenes. CNNs allow us to learn suitable feature representations for localization that are robust against motion blur and illumination changes. We make use of LSTM units on the CNN output, which play the role of a structured dimensionality reduction on the feature vector, leading to drastic improvements in localization performance. We provide extensive quantitative comparison of CNN-based and SIFT-based localization methods, showing the weaknesses and strengths of each. Furthermore, we present a new large-scale indoor dataset with accurate ground truth from a laser scanner. Experimental results on both indoor and outdoor public datasets show our method outperforms existing deep architectures, and can localize images in hard conditions, e.g., in the presence of mostly textureless surfaces, where classic SIFT-based methods fail.