Eduardo Valle
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Eduardo Valle has a total of 58 co-authors. Here are the top co-authors by number of shared publications:
- Sandra Avila
University of Campinas (UNICAMP)
8 shared publications - Alceu Bissoto
University of Bern
6 shared publications - Catarina Barata
Instituto Superior Técnico
4 shared publications - Pedro Tabacof
DCA / School of Electrical and Computer Engineering (FEEC) RECOD Lab University of Campinas (Unicamp) Campinas SP Brazil
4 shared publications Sandra Avila
4 shared publicationsAfonso Menegola
3 shared publications
Latest Publications
- 12 May 2025
- Published by Springer Nature
- p. 151-169
PAFUSE: Part-Based Diffusion for 3D Whole-Body Pose Estimation
- 12 May 2025
- Published by Springer Nature
- p. 151-169
PAFUSE: Part-Based Diffusion for 3D Whole-Body Pose Estimation
- Open Access
Even Small Correlation and Diversity Shifts Pose Dataset-Bias Issues
Distribution shifts are common in real-world datasets and can affect the performance and reliability of deep learning models. In this paper, we study two types of distribution shifts: diversity shifts, which occur when test samples exhibit patterns unseen during training, and correlation shifts, which occur when test data present a different correlation between seen invariant and spurious features. We propose an integrated protocol to analyze both types of shifts using datasets where they co-exist in a controllable manner. Finally, we apply our approach to a real-world classification problem of skin cancer analysis, using out-of-distribution datasets and specialized bias annotations. Our protocol reveals three findings: 1) Models learn and propagate correlation shifts even with low-bias training; this poses a risk of accumulating and combining unaccountable weak biases; 2) Models learn robust features in high- and low-bias scenarios but use spurious ones if test samples have them; this suggests that spurious correlations do not impair the learning of robust features; 3) Diversity shift can reduce the reliance on spurious correlations; this is counter intuitive since we expect biased models to depend more on biases when invariant features are missing. Our work has implications for distribution shift research and practice, providing new insights into how models learn and rely on spurious correlations under different types of shifts.
Open AccessEven Small Correlation and Diversity Shifts Pose Dataset-Bias Issues
Distribution shifts are common in real-world datasets and can affect the performance and reliability of deep learning models. In this paper, we study two types of distribution shifts: diversity shifts, which occur when test samples exhibit patterns unseen during training, and correlation shifts, which occur when test data present a different correlation between seen invariant and spurious features. We propose an integrated protocol to analyze both types of shifts using datasets where they co-exist in a controllable manner. Finally, we apply our approach to a real-world classification problem of skin cancer analysis, using out-of-distribution datasets and specialized bias annotations. Our protocol reveals three findings: 1) Models learn and propagate correlation shifts even with low-bias training; this poses a risk of accumulating and combining unaccountable weak biases; 2) Models learn robust features in high- and low-bias scenarios but use spurious ones if test samples have them; this suggests that spurious correlations do not impair the learning of robust features; 3) Diversity shift can reduce the reliance on spurious correlations; this is counter intuitive since we expect biased models to depend more on biases when invariant features are missing. Our work has implications for distribution shift research and practice, providing new insights into how models learn and rely on spurious correlations under different types of shifts.
- 29 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Adversarial Feature Hallucination in a Supervised Contrastive Space for Few-Shot Learning of Provenance in Paintings
Few-Shot Learning (FSL) is a problem formulation in which an intelligent system must learn discriminative patterns from a strictly reduced set of samples. While this can be achieved through many distinct strategies, Adversarial Feature Hallucination (AFH) has shown great promise when applied to general-purpose datasets. In this work, we expand AFH to the domain of artwork provenance analysis, and train a conditioned generative network to produce features from a given FSL context, which in turn are used to enrich the training of classifiers of few-sample data. We then further enrich the generative process by leveraging multi-objective optimization and Supervised Contrastive strategies to learn a more discriminative artwork feature space. We obtain encouraging results evaluated in an extensive painting dataset — considering authorship attribution —, and find that the hallucinated features can improve the decision accuracy by a substantial margin when compared to the baseline.
- 29 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Adversarial Feature Hallucination in a Supervised Contrastive Space for Few-Shot Learning of Provenance in Paintings
Few-Shot Learning (FSL) is a problem formulation in which an intelligent system must learn discriminative patterns from a strictly reduced set of samples. While this can be achieved through many distinct strategies, Adversarial Feature Hallucination (AFH) has shown great promise when applied to general-purpose datasets. In this work, we expand AFH to the domain of artwork provenance analysis, and train a conditioned generative network to produce features from a given FSL context, which in turn are used to enrich the training of classifiers of few-sample data. We then further enrich the generative process by leveraging multi-objective optimization and Supervised Contrastive strategies to learn a more discriminative artwork feature space. We obtain encouraging results evaluated in an extensive painting dataset — considering authorship attribution —, and find that the hallucinated features can improve the decision accuracy by a substantial margin when compared to the baseline.
- 14 October 2023
- Published by Springer Nature
- p. 105-114
The Performance of Transferability Metrics Does Not Translate to Medical Tasks
- 14 October 2023
- Published by Springer Nature
- p. 105-114
The Performance of Transferability Metrics Does Not Translate to Medical Tasks
- Top Cited Papers
- 10 June 2023
- Published by Elsevier in Medical Image Analysis
- Vol. 88, 102863
A survey on deep learning for skin lesion segmentation
- 10 June 2023
- Published by Elsevier in Medical Image Analysis
- Vol. 88, 102863
Top Cited PapersA survey on deep learning for skin lesion segmentation
- 14 February 2023
- Published by Springer Nature
- p. 133-149
Artifact-Based Domain Generalization of Skin Lesion Models
- 14 February 2023
- Published by Springer Nature
- p. 133-149
Artifact-Based Domain Generalization of Skin Lesion Models
- 14 February 2023
- Published by Springer Nature
- p. 150-166
An Evaluation of Self-supervised Pre-training for Skin-Lesion Analysis
- 14 February 2023
- Published by Springer Nature
- p. 150-166
An Evaluation of Self-supervised Pre-training for Skin-Lesion Analysis
- Open Access
An Evaluation of Self-Supervised Pre-Training for Skin-Lesion Analysis
Self-supervised pre-training appears as an advantageous alternative to supervised pre-trained for transfer learning. By synthesizing annotations on pretext tasks, self-supervision allows to pre-train models on large amounts of pseudo-labels before fine-tuning them on the target task. In this work, we assess self-supervision for the diagnosis of skin lesions, comparing three self-supervised pipelines to a challenging supervised baseline, on five test datasets comprising in- and out-of-distribution samples. Our results show that self-supervision is competitive both in improving accuracies and in reducing the variability of outcomes. Self-supervision proves particularly useful for low training data scenarios ($<1\,500$ and $<150$ samples), where its ability to stabilize the outcomes is essential to provide sound results.Open AccessAn Evaluation of Self-Supervised Pre-Training for Skin-Lesion Analysis
Self-supervised pre-training appears as an advantageous alternative to supervised pre-trained for transfer learning. By synthesizing annotations on pretext tasks, self-supervision allows to pre-train models on large amounts of pseudo-labels before fine-tuning them on the target task. In this work, we assess self-supervision for the diagnosis of skin lesions, comparing three self-supervised pipelines to a challenging supervised baseline, on five test datasets comprising in- and out-of-distribution samples. Our results show that self-supervision is competitive both in improving accuracies and in reducing the variability of outcomes. Self-supervision proves particularly useful for low training data scenarios ($<1\,500$ and $<150$ samples), where its ability to stabilize the outcomes is essential to provide sound results. - Open Access
Artifact-Based Domain Generalization of Skin Lesion Models
Deep Learning failure cases are abundant, particularly in the medical area. Recent studies in out-of-distribution generalization have advanced considerably on well-controlled synthetic datasets, but they do not represent medical imaging contexts. We propose a pipeline that relies on artifacts annotation to enable generalization evaluation and debiasing for the challenging skin lesion analysis context. First, we partition the data into levels of increasingly higher biased training and test sets for better generalization assessment. Then, we create environments based on skin lesion artifacts to enable domain generalization methods. Finally, after robust training, we perform a test-time debiasing procedure, reducing spurious features in inference images. Our experiments show our pipeline improves performance metrics in biased cases, and avoids artifacts when using explanation methods. Still, when evaluating such models in out-of-distribution data, they did not prefer clinically-meaningful features. Instead, performance only improved in test sets that present similar artifacts from training, suggesting models learned to ignore the known set of artifacts. Our results raise a concern that debiasing models towards a single aspect may not be enough for fair skin lesion analysis.
Open AccessArtifact-Based Domain Generalization of Skin Lesion Models
Deep Learning failure cases are abundant, particularly in the medical area. Recent studies in out-of-distribution generalization have advanced considerably on well-controlled synthetic datasets, but they do not represent medical imaging contexts. We propose a pipeline that relies on artifacts annotation to enable generalization evaluation and debiasing for the challenging skin lesion analysis context. First, we partition the data into levels of increasingly higher biased training and test sets for better generalization assessment. Then, we create environments based on skin lesion artifacts to enable domain generalization methods. Finally, after robust training, we perform a test-time debiasing procedure, reducing spurious features in inference images. Our experiments show our pipeline improves performance metrics in biased cases, and avoids artifacts when using explanation methods. Still, when evaluating such models in out-of-distribution data, they did not prefer clinically-meaningful features. Instead, performance only improved in test sets that present similar artifacts from training, suggesting models learned to ignore the known set of artifacts. Our results raise a concern that debiasing models towards a single aspect may not be enough for fair skin lesion analysis.
- Open Access
PODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning
Lifelong learning has attracted much attention, but existing works still struggle to fight catastrophic forgetting and accumulate knowledge over long stretches of incremental learning. In this work, we propose PODNet, a model inspired by representation learning. By carefully balancing the compromise between remembering the old classes and learning new ones, PODNet fights catastrophic forgetting, even over very long runs of small incremental tasks --a setting so far unexplored by current works. PODNet innovates on existing art with an efficient spatial-based distillation-loss applied throughout the model and a representation comprising multiple proxy vectors for each class. We validate those innovations thoroughly, comparing PODNet with three state-of-the-art models on three datasets: CIFAR100, ImageNet100, and ImageNet1000. Our results showcase a significant advantage of PODNet over existing art, with accuracy gains of 12.10, 6.51, and 2.85 percentage points, respectively. Code is available at https://github.com/arthurdouillard/incremental_learning.pytorch
Open AccessPODNet: Pooled Outputs Distillation for Small-Tasks Incremental Learning
Lifelong learning has attracted much attention, but existing works still struggle to fight catastrophic forgetting and accumulate knowledge over long stretches of incremental learning. In this work, we propose PODNet, a model inspired by representation learning. By carefully balancing the compromise between remembering the old classes and learning new ones, PODNet fights catastrophic forgetting, even over very long runs of small incremental tasks --a setting so far unexplored by current works. PODNet innovates on existing art with an efficient spatial-based distillation-loss applied throughout the model and a representation comprising multiple proxy vectors for each class. We validate those innovations thoroughly, comparing PODNet with three state-of-the-art models on three datasets: CIFAR100, ImageNet100, and ImageNet1000. Our results showcase a significant advantage of PODNet over existing art, with accuracy gains of 12.10, 6.51, and 2.85 percentage points, respectively. Code is available at https://github.com/arthurdouillard/incremental_learning.pytorch