Search

Continual learning aims to emulate the human ability to continually accumulate knowledge over sequential tasks. The main challenge is to maintain performance on previously learned tasks after learning new tasks, i.e., to avoid catastrophic forgetting. We propose a Channel-wise Lightweight Reprogramming (CLR) approach that helps convolutional neural networks (CNNs) overcome catastrophic forgetting during continual learning. We show that a CNN model trained on an old task (or self-supervised proxy task) could be ``reprogrammed" to solve a new task by using our proposed lightweight (very cheap) reprogramming parameter. With the help of CLR, we have a better stability-plasticity trade-off to solve continual learning problems: To maintain stability and retain previous task ability, we use a common task-agnostic immutable part as the shared ``anchor" parameter set. We then add task-specific lightweight reprogramming parameters to reinterpret the outputs of the immutable parts, to enable plasticity and integrate new knowledge. To learn sequential tasks, we only train the lightweight reprogramming parameters to learn each new task. Reprogramming parameters are task-specific and exclusive to each task, which makes our method immune to catastrophic forgetting. To minimize the parameter requirement of reprogramming to learn new tasks, we make reprogramming lightweight by only adjusting essential kernels and learning channel-wise linear mappings from anchor parameters to task-specific domain knowledge. We show that, for general CNNs, the CLR parameter increase is less than 0.6\% for any new task. Our method outperforms 13 state-of-the-art continual learning baselines on a new challenging sequence of 53 image classification datasets. Code and data are available at https://github.com/gyhandy/Channel-wise-Lightweight-Reprogramming, ICCV 2023

We focus on the challenge of out-of-distribution (OOD) detection in deep learning models, a crucial aspect in ensuring reliability. Despite considerable effort, the problem remains significantly challenging in deep learning models due to their propensity to output over-confident predictions for OOD inputs. We propose a novel one-class open-set OOD detector that leverages text-image pre-trained models in a zero-shot fashion and incorporates various descriptions of in-domain and OOD. Our approach is designed to detect anything not in-domain and offers the flexibility to detect a wide variety of OOD, defined via fine- or coarse-grained labels, or even in natural language. We evaluate our approach on challenging benchmarks including large-scale datasets containing fine-grained, semantically similar classes, distributionally shifted images, and multi-object images containing a mixture of in-domain and OOD objects. Our method shows superior performance over previous methods on all benchmarks. Code is available at https://github.com/gyhandy/One-Class-Anything, Comment: 16 pages (including appendix and references), 3 figures

Ophiostomatoid fungi exhibit a complex relationship with bark beetles; exhausting of host tree defenses is traditionally regarded as one of the key benefits provided to beetle vectors. Ophiostoma bicolor is one of the dominant species of the mycobiota associated with Ips genus bark beetles which infect the spruce trees across the Eurasian continent. Host spruce trees resist fungal invasion through structural and inducible defenses, but the underlying mechanisms at the molecular level, particularly with respect to the interaction between bark beetle-associated fungi and host trees, remain unclear. The aim of this study was to observe the pathological physiology and molecular changes in Picea koraiensis seedlings after artificial inoculation with O. bicolor strains (TS, BH, QH, MX, and LWQ). This study showed that O. bicolor was a weakly virulent pathogen of spruce, and that the virulent of the five O. bicolor strains showed differentiation. All O. bicolor strains could induce monoterpenoid release. A positive correlation between fungal virulence and release of monoterpenoids was observed. Furthermore, the release rate of monoterpenoids peaked at 4 days post-inoculation (dpi) and then decreased from 4 to 90 dpi. Transcriptomic analysis at 4 dpi showed that many plant-pathogen interaction processes and mitogen-activated protein kinase (MAPK) metabolic processes were activated. The expression of monoterpenoid precursor synthesis genes and diterpenoid synthesis genes was upregulated, indicating that gene expression regulated the release rate of monoterpenoids at 4 dpi. The enriched pathways may reveal the immune response mechanism of spruce to ophiostomatoid fungi. The dominant O. bicolor possibly induces the host defense rather than defense depletion, which is likely the pattern conducted by the pioneers of beetle-associated mycobiota, such as Endoconidiophora spp.. Overall, these results facilitate a better understanding of the interaction mechanism between the dominant association of beetles and the host at the molecular level.

Object pose increases intraclass object variance which makes object recognition from 2D images harder. To render a classifier robust to pose variations, most deep neural networks try to eliminate the influence of pose by using large datasets with many poses for each class. Here, we propose a different approach: a class-agnostic object pose transformation network (OPT-Net) can transform an image along 3D yaw and pitch axes to synthesize additional poses continuously. Synthesized images lead to better training of an object classifier. We design a novel eliminate-add structure to explicitly disentangle pose from object identity: first eliminate pose information of the input image and then add target pose information (regularized as continuous variables) to synthesize any target pose. We trained OPT-Net on images of toy vehicles shot on a turntable from the iLab-20M dataset. After training on unbalanced discrete poses (5 classes with 6 poses per object instance, plus 5 classes with only 2 poses), we show that OPT-Net can synthesize balanced continuous new poses along yaw and pitch axes with high quality. Training a ResNet-18 classifier with original plus synthesized poses improves mAP accuracy by 9% overtraining on original poses only. Further, the pre-trained OPT-Net can generalize to new object classes, which we demonstrate on both iLab-20M and RGB-D. We also show that the learned features can generalize to ImageNet., Comment: ECCV 2020, with supplementary materials

We propose to learn multiple local Mahalanobis distance metrics to perform k-nearest neighbor (kNN) classification of temporal sequences. Temporal sequences are first aligned by dynamic time warping (DTW); given the alignment path, similarity between two sequences is measured by the DTW distance, which is computed as the accumulated distance between matched temporal point pairs along the alignment path. Traditionally, Euclidean metric is used for distance computation between matched pairs, which ignores the data regularities and might not be optimal for applications at hand. Here we propose to learn multiple Mahalanobis metrics, such that DTW distance becomes the sum of Mahalanobis distances. We adapt the large margin nearest neighbor (LMNN) framework to our case, and formulate multiple metric learning as a linear programming problem. Extensive sequence classification results show that our proposed multiple metrics learning approach is effective, insensitive to the preceding alignment qualities, and reaches the state-of-the-art performances on UCR time series datasets.

Continual Learning in artificial neural networks suffers from interference and forgetting when different tasks are learned sequentially. This paper introduces the Active Long Term Memory Networks (A-LTM), a model of sequential multi-task deep learning that is able to maintain previously learned association between sensory input and behavioral output while acquiring knew knowledge. A-LTM exploits the non-convex nature of deep neural networks and actively maintains knowledge of previously learned, inactive tasks using a distillation loss. Distortions of the learned input-output map are penalized but hidden layers are free to transverse towards new local optima that are more favorable for the multi-task objective. We re-frame the McClelland's seminal Hippocampal theory with respect to Catastrophic Inference (CI) behavior exhibited by modern deep architectures trained with back-propagation and inhomogeneous sampling of latent factors across epochs. We present empirical results of non-trivial CI during continual learning in Deep Linear Networks trained on the same task, in Convolutional Neural Networks when the task shifts from predicting semantic to graphical factors and during domain adaptation from simple to complex environments. We present results of the A-LTM model's ability to maintain viewpoint recognition learned in the highly controlled iLab-20M dataset with 10 object categories and 88 camera viewpoints, while adapting to the unstructured domain of Imagenet with 1,000 object categories.

The chemical constituents in the roots of Aristolochia liukiuensis Hatsusima were examined. Aristolactone (1), mansonone G (3), dehydrooxoperezinone (4), aristololactam D2 (6), 3,4 - methylenedioxy - 8 - methoxyphenanthrene - 1 - carboxylic acid (8), aristolochic acid 2 (10), and aristolochic acid 4 methyl ester (11) were isolated from the roots, in addition to already isolated constituents from the leaves and the stems, such as sitosterol (2), 2-hydroxy-1-methoxy-4,5-dioxo-6,7-dehydroaporphine (5), aristololactam A2 (7), aristolochic acid 1 (9), and sitosterolfl-β-glucoside (12). This is the first report of 4 in natural sources., 紀要論文