Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userIt Takes Two to Tango: Mixup for Deep Metric Learning
127
0
0.0
(
0
)
Added by
Shashanka Venkataramanan
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Metric learning involves learning a discriminative representation such that embeddings of similar classes are encouraged to be close, while embeddings of dissimilar classes are pushed far apart. State-of-the-art methods focus mostly on sophisticated loss functions or mining strategies. On the one hand, metric learning losses consider two or more examples at a time. On the other hand, modern data augmentation methods for classification consider two or more examples at a time. The combination of the two ideas is under-studied. In this work, we aim to bridge this gap and improve representations using mixup, which is a powerful data augmentation approach interpolating two or more examples and corresponding target labels at a time. This task is challenging because, unlike classification, the loss functions used in metric learning are not additive over examples, so the idea of interpolating target labels is not straightforward. To the best of our knowledge, we are the first to investigate mixing examples and target labels for deep metric learning. We develop a generalized formulation that encompasses existing metric learning loss functions and modify it to accommodate for mixup, introducing Metric Mix, or Metrix. We show that mixing inputs, intermediate representations or embeddings along with target labels significantly improves representations and outperforms state-of-the-art metric learning methods on four benchmark datasets.
rate research
Read More
K2-146 is a cool, 0.358 M_sun dwarf that was found to host a mini-Neptune with a 2.67-days period. The planet exhibited strong transit timing variations (TTVs) of greater than 30 minutes, indicative of the presence of a further object in the system. Here we report the discovery of the previously undetected outer planet, K2-146 c, in the system using additional photometric data. K2-146 c was found to have a grazing transit geometry and a 3.97-day period. The outer planet was only significantly detected in the latter K2 campaigns presumably because of precession of its orbital plane. The TTVs of K2-146 b and c were measured using observations spanning a baseline of almost 1200 days. We found strong anti-correlation in the TTVs, suggesting the two planets are gravitationally interacting. Our TTV and transit model analyses revealed that K2-146 b has a radius of 2.25 $pm$ 0.10 R_earth and a mass of 5.6 $pm$ 0.7 M_earth, whereas K2-146 c has a radius of $2.59_{-0.39}^{+1.81}$ R_earth and a mass of 7.1 $pm$ 0.9 M_earth. The inner and outer planets likely have moderate eccentricities of $e = 0.14 pm 0.07$ and $0.16 pm 0.07$, respectively. Long-term numerical integrations of the two-planet orbital solution show that it can be dynamically stable for at least 2 Myr. The evaluation of the resonance angles of the planet pair indicates that K2-146 b and c are likely trapped in a 3:2 mean motion resonance. The orbital architecture of the system points to a possible convergent migration origin.
Recent studies have revealed that neural network-based policies can be easily fooled by adversarial examples. However, while most prior works analyze the effects of perturbing every pixel of every frame assuming white-box policy access, in this paper we take a more restrictive view towards adversary generation - with the goal of unveiling the limits of a models vulnerability. In particular, we explore minimalistic attacks by defining three key settings: (1) black-box policy access: where the attacker only has access to the input (state) and output (action probability) of an RL policy; (2) fractional-state adversary: where only several pixels are perturbed, with the extreme case being a single-pixel adversary; and (3) tactically-chanced attack: where only significant frames are tactically chosen to be attacked. We formulate the adversarial attack by accommodating the three key settings and explore their potency on six Atari games by examining four fully trained state-of-the-art policies. In Breakout, for example, we surprisingly find that: (i) all policies showcase significant performance degradation by merely modifying 0.01% of the input state, and (ii) the policy trained by DQN is totally deceived by perturbation to only 1% frames.
We describe our system for SemEval-2020 Task 11 on Detection of Propaganda Techniques in News Articles. We developed ensemble models using RoBERTa-based neural architectures, additional CRF layers, transfer learning between the two subtasks, and advanced post-processing to handle the multi-label nature of the task, the consistency between nested spans, repetitions, and labels from similar spans in training. We achieved sizable improvements over baseline fine-tuned RoBERTa models, and the official evaluation ranked our system 3rd (almost tied with the 2nd) out of 36 teams on the span identification subtask with an F1 score of 0.491, and 2nd (almost tied with the 1st) out of 31 teams on the technique classification subtask with an F1 score of 0.62.
It has been widely recognized that adversarial examples can be easily crafted to fool deep networks, which mainly root from the locally non-linear behavior nearby input examples. Applying mixup in training provides an effective mechanism to improve generalization performance and model robustness against adversarial perturbations, which introduces the globally linear behavior in-between training examples. However, in previous work, the mixup-trained models only passively defend adversarial attacks in inference by directly classifying the inputs, where the induced global linearity is not well exploited. Namely, since the locality of the adversarial perturbations, it would be more efficient to actively break the locality via the globality of the model predictions. Inspired by simple geometric intuition, we develop an inference principle, named mixup inference (MI), for mixup-trained models. MI mixups the input with other random clean samples, which can shrink and transfer the equivalent perturbation if the input is adversarial. Our experiments on CIFAR-10 and CIFAR-100 demonstrate that MI can further improve the adversarial robustness for the models trained by mixup and its variants.
Distance Metric Learning (DML) seeks to learn a discriminative embedding where similar examples are closer, and dissimilar examples are apart. In this paper, we address the problem of Semi-Supervised DML (SSDML) that tries to learn a metric using a few labeled examples, and abundantly available unlabeled examples. SSDML is important because it is infeasible to manually annotate all the examples present in a large dataset. Surprisingly, with the exception of a few classical approaches that learn a linear Mahalanobis metric, SSDML has not been studied in the recent years, and lacks approaches in the deep SSDML scenario. In this paper, we address this challenging problem, and revamp SSDML with respect to deep learning. In particular, we propose a stochastic, graph-based approach that first propagates the affinities between the pairs of examples from labeled data, to that of the unlabeled pairs. The propagated affinities are used to mine triplet based constraints for metric learning. We impose orthogonality constraint on the metric parameters, as it leads to a better performance by avoiding a model collapse.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
