Recent work studies the supervised online continual learning setting where a learner receives a stream of data whose class distribution changes over time. Distinct from other continual learning settings the learner is presented new samples only once and must distinguish between all seen classes. A number of successful methods in this setting focus on storing and replaying a subset of samples alongside incoming data in a computationally efficient manner. One recent proposal ER-AML achieved strong performance in this setting by applying an asymmetric loss based on contrastive learning to the incoming data and replayed data. However, a key ingredient of the proposed method is avoiding contrasts between incoming data and stored data, which makes it impractical for the setting where only one new class is introduced in each phase of the stream. In this work we adapt a recently proposed approach (\textit{BYOL}) from self-supervised learning to the supervised learning setting, unlocking the constraint on contrasts. We then show that supplementing this with additional regularization on class prototypes yields a new method that achieves strong performance in the one-class incremental learning setting and is competitive with the top performing methods in the multi-class incremental setting.
In the online continual learning paradigm, agents must learn from a changing distribution while respecting memory and compute constraints. Experience Replay (ER), where a small subset of past data is stored and replayed alongside new data, has emerged as a simple and effective learning strategy. In this work, we focus on the change in representations of observed data that arises when previously unobserved classes appear in the incoming data stream, and new classes must be distinguished from previous ones. We shed new light on this question by showing that applying ER causes the newly added classes' representations to overlap significantly with the previous classes, leading to highly disruptive parameter updates. Based on this empirical analysis, we propose a new method which mitigates this issue by shielding the learned representations from drastic adaptation to accommodate new classes. We show that using an asymmetric update rule pushes new classes to adapt to the older ones (rather than the reverse), which is more effective especially at task boundaries, where much of the forgetting typically occurs. Empirical results show significant gains over strong baselines on standard continual learning benchmarks
The development of biologically-plausible learning algorithms is important for understanding learning in the brain, but most of them fail to scale-up to real-world tasks, limiting their potential as explanations for learning by real brains. As such, it is important to explore learning algorithms that come with strong theoretical guarantees and can match the performance of backpropagation (BP) on complex tasks. One such algorithm is Difference Target Propagation (DTP), a biologically-plausible learning algorithm whose close relation with Gauss-Newton (GN) optimization has been recently established. However, the conditions under which this connection rigorously holds preclude layer-wise training of the feedback pathway synaptic weights (which is more biologically plausible). Moreover, good alignment between DTP weight updates and loss gradients is only loosely guaranteed and under very specific conditions for the architecture being trained. In this paper, we propose a novel feedback weight training scheme that ensures both that DTP approximates BP and that layer-wise feedback weight training can be restored without sacrificing any theoretical guarantees. Our theory is corroborated by experimental results and we report the best performance ever achieved by DTP on CIFAR-10 and ImageNet 32$\times$32
Federated learning is an emerging paradigm that permits a large number of clients with heterogeneous data to coordinate learning of a unified global model without the need to share data amongst each other. Standard federated learning algorithms involve averaging of model parameters or gradient updates to approximate the global model at the server. However, in heterogeneous settings averaging can result in information loss and lead to poor generalization due to the bias induced by dominant clients. We hypothesize that to generalize better across non-i.i.d datasets as in FL settings, the algorithms should focus on learning the invariant mechanism that is constant while ignoring spurious mechanisms that differ across clients. Inspired from recent work in the Out-of-Distribution literature, we propose a gradient masked averaging approach for federated learning as an alternative to the standard averaging of client updates. This client update aggregation technique can be adapted as a drop-in replacement in most existing federated algorithms. We perform extensive experiments with gradient masked approach on multiple FL algorithms with in-distribution, real-world, and out-of-distribution (as the worst case scenario) test dataset and show that it provides consistent improvements, particularly in the case of heterogeneous clients.
The wavelet scattering transform creates geometric invariants and deformation stability from an initial structured signal. In multiple signal domains it has been shown to yield more discriminative representations compared to other non-learned representations, and to outperform learned representations in certain tasks, particularly on limited labeled data and highly structured signals. The wavelet filters used in the scattering transform are typically selected to create a tight frame via a parameterized mother wavelet. Focusing on Morlet wavelets, we propose to instead adapt the scales, orientations, and slants of the filters to produce problem-specific parametrizations of the scattering transform. We show that our learned versions of the scattering transform yield significant performance gains over the standard scattering transform in the small sample classification settings, and our empirical results suggest that tight frames may not always be necessary for scattering transforms to extract effective representations.
The impressive performance of deep convolutional neural networks in single-view 3D reconstruction suggests that these models perform non-trivial reasoning about the 3D structure of the output space. Recent work has challenged this belief, showing that, on standard benchmarks, complex encoder-decoder architectures perform similarly to nearest-neighbor baselines or simple linear decoder models that exploit large amounts of per-category data. However, building large collections of 3D shapes for supervised training is a laborious process; a more realistic and less constraining task is inferring 3D shapes for categories with few available training examples, calling for a model that can successfully generalize to novel object classes. In this work we experimentally demonstrate that naive baselines fail in this few-shot learning setting, in which the network must learn informative shape priors for inference of new categories. We propose three ways to learn a class-specific global shape prior, directly from data. Using these techniques, we are able to capture multi-scale information about the 3D shape, and account for intra-class variability by virtue of an implicit compositional structure. Experiments on the popular ShapeNet dataset show that our method outperforms a zero-shot baseline by over 40%, and the current state-of-the-art by over 10%, in terms of relative performance, in the few-shot setting.
A commonly cited inefficiency of neural network training using back-propagation is the update locking problem: each layer must wait for the signal to propagate through the full network before updating. Several alternatives that can alleviate this issue have been proposed. In this context, we consider a simple alternative based on minimal feedback, which we call Decoupled Greedy Learning (DGL). It is based on a classic greedy relaxation of the joint training objective, recently shown to be effective in the context of Convolutional Neural Networks (CNNs) on large-scale image classification. We consider an optimization of this objective that permits us to decouple the layer training, allowing for layers or modules in networks to be trained with a potentially linear parallelization. With the use of a replay buffer we show that this approach can be extended to asynchronous settings, where modules can operate and continue to update with possibly large communication delays. To address bandwidth and memory issues we propose an approach based on online vector quantization. This allows to drastically reduce the communication bandwidth between modules and required memory for replay buffers. We show theoretically and empirically that this approach converges and compare it to the sequential solvers. We demonstrate the effectiveness of DGL against alternative approaches on the CIFAR-10 dataset and on the large-scale ImageNet dataset.
We study the online continual learning paradigm, where agents must learn from a changing distribution with constrained memory and compute. Previous work often tackle catastrophic forgetting by overcoming changes in the space of model parameters. In this work we instead focus on the change in representations of previously observed data due to the introduction of previously unobserved class samples in the incoming data stream. We highlight the issues that arise in the practical setting where new classes must be distinguished between all previous classes. Starting from a popular approach, experience replay, we consider a metric learning based loss function, the triplet loss, which allows us to more explicitly constrain the behavior of representations. We hypothesize and empirically confirm that the selection of negatives used in the triplet loss plays a major role in the representation change, or drift, of previously observed data and can be greatly reduced by appropriate negative selection. Motivated by this we further introduce a simple adjustment to the standard cross entropy loss used in prior experience replay that achieves similar effect. Our approach greatly improves the performance of experience replay and obtains state-of-the-art on several existing benchmarks in online continual learning, while remaining efficient in both memory and compute.
A recent line of work showed that various forms of convolutional kernel methods can be competitive with standard supervised deep convolutional networks on datasets like CIFAR-10, obtaining accuracies in the range of 87-90% while being more amenable to theoretical analysis. In this work, we highlight the importance of a data-dependent feature extraction step that is key to the obtain good performance in convolutional kernel methods. This step typically corresponds to a whitened dictionary of patches, and gives rise to a data-driven convolutional kernel methods. We extensively study its effect, demonstrating it is the key ingredient for high performance of these methods. Specifically, we show that one of the simplest instances of such kernel methods, based on a single layer of image patches followed by a linear classifier is already obtaining classification accuracies on CIFAR-10 in the same range as previous more sophisticated convolutional kernel methods. We scale this method to the challenging ImageNet dataset, showing such a simple approach can exceed all existing non-learned representation methods. This is a new baseline for object recognition without representation learning methods, that initiates the investigation of convolutional kernel models on ImageNet. We conduct experiments to analyze the dictionary that we used, our ablations showing they exhibit low-dimensional properties.