LadaBERT: Lightweight Adaptation of BERT through Hybrid Model Compression

BERT is a cutting-edge language representation model pre-trained by a large corpus, which achieves superior performances on various natural language understanding tasks. However, a major blocking issue of applying BERT to online services is that it is memory-intensive and leads to unsatisfactory latency of user requests, raising the necessity of model compression. Existing solutions leverage the knowledge distillation framework to learn a smaller model that imitates the behaviors of BERT. However, the training procedure of knowledge distillation is expensive itself as it requires sufficient training data to imitate the teacher model. In this paper, we address this issue by proposing a hybrid solution named LadaBERT (Lightweight adaptation of BERT through hybrid model compression), which combines the advantages of different model compression methods, including weight pruning, matrix factorization and knowledge distillation. LadaBERT achieves state-of-the-art accuracy on various public datasets while the training overheads can be reduced by an order of magnitude.

Improving BERT with Self-Supervised Attention

One of the most popular paradigms of applying large, pre-trained NLP models such as BERT is to fine-tune it on a smaller dataset. However, one challenge remains as the fine-tuned model often overfits on smaller datasets. A symptom of this phenomenon is that irrelevant words in the sentences, even when they are obvious to humans, can substantially degrade the performance of these fine-tuned BERT models. In this paper, we propose a novel technique, called Self-Supervised Attention (SSA) to help facilitate this generalization challenge. Specifically, SSA automatically generates weak, token-level attention labels iteratively by "probing" the fine-tuned model from the previous iteration. We investigate two different ways of integrating SSA into BERT and propose a hybrid approach to combine their benefits. Empirically, on a variety of public datasets, we illustrate significant performance improvement using our SSA-enhanced BERT model.

Semantic Flow for Fast and Accurate Scene Parsing

In this paper, we focus on effective methods for fast and accurate scene parsing. A common practice to improve the performance is to attain high resolution feature maps with strong semantic representation. Two strategies are widely used---astrous convolutions and feature pyramid fusion, are either computation intensive or ineffective. Inspired by Optical Flow for motion alignment between adjacent video frames, we propose a Flow Alignment Module (FAM) to learn Semantic Flow between feature maps of adjacent levels and broadcast high-level features to high resolution features effectively and efficiently. Furthermore, integrating our module to a common feature pyramid structure exhibits superior performance over other real-time methods even on very light-weight backbone networks, such as ResNet-18. Extensive experiments are conducted on several challenging datasets, including Cityscapes, PASCAL Context, ADE20K and CamVid. Particularly, our network is the first to achieve 80.4\% mIoU on Cityscapes with a frame rate of 26 FPS. The code will be available at \url{https://github.com/donnyyou/torchcv}.

TextNAS: A Neural Architecture Search Space tailored for Text Representation

Learning text representation is crucial for text classification and other language related tasks. There are a diverse set of text representation networks in the literature, and how to find the optimal one is a non-trivial problem. Recently, the emerging Neural Architecture Search (NAS) techniques have demonstrated good potential to solve the problem. Nevertheless, most of the existing works of NAS focus on the search algorithms and pay little attention to the search space. In this paper, we argue that the search space is also an important human prior to the success of NAS in different applications. Thus, we propose a novel search space tailored for text representation. Through automatic search, the discovered network architecture outperforms state-of-the-art models on various public datasets on text classification and natural language inference tasks. Furthermore, some of the design principles found in the automatic network agree well with human intuition.

Customized Graph Embedding: Tailoring the Embedding Vector to a Specific Application

The graph is a natural representation of data in a variety of real-world applications, for example as a knowledge graph, a social network, or a biological network. To better leverage the information behind the data, the method of graph embedding is recently proposed and extensively studied. The traditional graph embedding method, while it provides an effective way to understand what is behind the graph data, is unfortunately sub-optimal in many cases. This is because its learning procedure is disconnected from the target application. In this paper, we propose a novel approach, Customized Graph Embedding (CGE), to tackle this problem. The CGE algorithm learns a customized vector representation of the graph by differentiating the varying importance of distinct graph paths. Experiments are carried out on a diverse set of node classification datasets and strong performance is demonstrated.

Global Aggregation then Local Distribution in Fully Convolutional Networks

It has been widely proven that modelling long-range dependencies in fully convolutional networks (FCNs) via global aggregation modules is critical for complex scene understanding tasks such as semantic segmentation and object detection. However, global aggregation is often dominated by features of large patterns and tends to oversmooth regions that contain small patterns (e.g., boundaries and small objects). To resolve this problem, we propose to first use \emph{Global Aggregation} and then \emph{Local Distribution}, which is called GALD, where long-range dependencies are more confidently used inside large pattern regions and vice versa. The size of each pattern at each position is estimated in the network as a per-channel mask map. GALD is end-to-end trainable and can be easily plugged into existing FCNs with various global aggregation modules for a wide range of vision tasks, and consistently improves the performance of state-of-the-art object detection and instance segmentation approaches. In particular, GALD used in semantic segmentation achieves new state-of-the-art performance on Cityscapes test set with mIoU 83.3\%. Code is available at: \url{https://github.com/lxtGH/GALD-Net}

Dual Graph Convolutional Network for Semantic Segmentation

Exploiting long-range contextual information is key for pixel-wise prediction tasks such as semantic segmentation. In contrast to previous work that uses multi-scale feature fusion or dilated convolutions, we propose a novel graph-convolutional network (GCN) to address this problem. Our Dual Graph Convolutional Network (DGCNet) models the global context of the input feature by modelling two orthogonal graphs in a single framework. The first component models spatial relationships between pixels in the image, whilst the second models interdependencies along the channel dimensions of the network's feature map. This is done efficiently by projecting the feature into a new, lower-dimensional space where all pairwise interactions can be modelled, before reprojecting into the original space. Our simple method provides substantial benefits over a strong baseline and achieves state-of-the-art results on both Cityscapes (82.0\% mean IoU) and Pascal Context (53.7\% mean IoU) datasets.

Reprojection R-CNN: A Fast and Accurate Object Detector for 360° Images

360{\deg} images are usually represented in either equirectangular projection (ERP) or multiple perspective projections. Different from the flat 2D images, the detection task is challenging for 360{\deg} images due to the distortion of ERP and the inefficiency of perspective projections. However, existing methods mostly focus on one of the above representations instead of both, leading to limited detection performance. Moreover, the lack of appropriate bounding-box annotations as well as the annotated datasets further increases the difficulties of the detection task. In this paper, we present a standard object detection framework for 360{\deg} images. Specifically, we adapt the terminologies of the traditional object detection task to the omnidirectional scenarios, and propose a novel two-stage object detector, i.e., Reprojection R-CNN by combining both ERP and perspective projection. Owing to the omnidirectional field-of-view of ERP, Reprojection R-CNN first generates coarse region proposals efficiently by a distortion-aware spherical region proposal network. Then, it leverages the distortion-free perspective projection and refines the proposed regions by a novel reprojection network. We construct two novel synthetic datasets for training and evaluation. Experiments reveal that Reprojection R-CNN outperforms the previous state-of-the-art methods on the mAP metric. In addition, the proposed detector could run at 178ms per image in the panoramic datasets, which implies its practicability in real-world applications.

GFF: Gated Fully Fusion for Semantic Segmentation

Semantic segmentation generates comprehensive understanding of scenes at a semantic level through densely predicting the category for each pixel. High-level features from Deep Convolutional Neural Networks already demonstrate their effectiveness in semantic segmentation tasks, however the coarse resolution of high-level features often leads to inferior results for small/thin objects where detailed information is important but missing. It is natural to consider importing low level features to compensate the lost detailed information in high level representations. Unfortunately, simply combining multi-level features is less effective due to the semantic gap existing among them. In this paper, we propose a new architecture, named Gated Fully Fusion(GFF), to selectively fuse features from multiple levels using gates in a fully connected way. Specifically, features at each level are enhanced by higher-level features with stronger semantics and lower-level features with more details, and gates are used to control the propagation of useful information which significantly reduces the noises during fusion. We achieve the state of the art results on two challenging scene understanding datasets, i.e., 82.3% mIoU on Cityscapes test set and 45.3% mIoU on ADE20K validation set. Codes and the trained models will be made publicly available.

A Cooperative Multi-Agent Reinforcement Learning Framework for Resource Balancing in Complex Logistics Network

Resource balancing within complex transportation networks is one of the most important problems in real logistics domain. Traditional solutions on these problems leverage combinatorial optimization with demand and supply forecasting. However, the high complexity of transportation routes, severe uncertainty of future demand and supply, together with non-convex business constraints make it extremely challenging in the traditional resource management field. In this paper, we propose a novel sophisticated multi-agent reinforcement learning approach to address these challenges. In particular, inspired by the externalities especially the interactions among resource agents, we introduce an innovative cooperative mechanism for state and reward design resulting in more effective and efficient transportation. Extensive experiments on a simulated ocean transportation service demonstrate that our new approach can stimulate cooperation among agents and lead to much better performance. Compared with traditional solutions based on combinatorial optimization, our approach can give rise to a significant improvement in terms of both performance and stability.