We present a new dataset with the goal of advancing the state-of-the-art in object recognition by placing the question of object recognition in the context of the broader question of scene understanding. This is achieved by gathering images of complex everyday scenes containing common objects in their natural context. Objects are labeled using per-instance segmentations to aid in precise object localization. Our dataset contains photos of 91 objects types that would be easily recognizable by a 4 year old. With a total of 2.5 million labeled instances in 328k images, the creation of our dataset drew upon extensive crowd worker involvement via novel user interfaces for category detection, instance spotting and instance segmentation. We present a detailed statistical analysis of the dataset in comparison to PASCAL, ImageNet, and SUN. Finally, we provide baseline performance analysis for bounding box and segmentation detection results using a Deformable Parts Model.

/pdf/microsoft-coco-common-objects-in-context-55ihxq1zt5.pdf

Microsoft COCO: Common Objects in Context

We introduce a new language representation model called BERT, which stands for Bidirectional Encoder Representations from Transformers. Unlike recent language representation models, BERT is designed to pre-train deep bidirectional representations from unlabeled text by jointly conditioning on both left and right context in all layers. As a result, the pre-trained BERT model can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of tasks, such as question answering and language inference, without substantial task-specific architecture modifications. 
BERT is conceptually simple and empirically powerful. It obtains new state-of-the-art results on eleven natural language processing tasks, including pushing the GLUE score to 80.5% (7.7% point absolute improvement), MultiNLI accuracy to 86.7% (4.6% absolute improvement), SQuAD v1.1 question answering Test F1 to 93.2 (1.5 point absolute improvement) and SQuAD v2.0 Test F1 to 83.1 (5.1 point absolute improvement).

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

We introduce a new language representation model called BERT, which stands for Bidirectional Encoder Representations from Transformers. Unlike recent language representation models (Peters et al., 2018a; Radford et al., 2018), BERT is designed to pre-train deep bidirectional representations from unlabeled text by jointly conditioning on both left and right context in all layers. As a result, the pre-trained BERT model can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of tasks, such as question answering and language inference, without substantial task-specific architecture modifications. BERT is conceptually simple and empirically powerful. It obtains new state-of-the-art results on eleven natural language processing tasks, including pushing the GLUE score to 80.5 (7.7 point absolute improvement), MultiNLI accuracy to 86.7% (4.6% absolute improvement), SQuAD v1.1 question answering Test F1 to 93.2 (1.5 point absolute improvement) and SQuAD v2.0 Test F1 to 83.1 (5.1 point absolute improvement).

We present a method for detecting objects in images using a single deep neural network. Our approach, named SSD, discretizes the output space of bounding boxes into a set of default boxes over different aspect ratios and scales per feature map location. At prediction time, the network generates scores for the presence of each object category in each default box and produces adjustments to the box to better match the object shape. Additionally, the network combines predictions from multiple feature maps with different resolutions to naturally handle objects of various sizes. SSD is simple relative to methods that require object proposals because it completely eliminates proposal generation and subsequent pixel or feature resampling stages and encapsulates all computation in a single network. This makes SSD easy to train and straightforward to integrate into systems that require a detection component. Experimental results on the PASCAL VOC, COCO, and ILSVRC datasets confirm that SSD has competitive accuracy to methods that utilize an additional object proposal step and is much faster, while providing a unified framework for both training and inference. For \(300 \times 300\) input, SSD achieves 74.3 % mAP on VOC2007 test at 59 FPS on a Nvidia Titan X and for \(512 \times 512\) input, SSD achieves 76.9 % mAP, outperforming a comparable state of the art Faster R-CNN model. Compared to other single stage methods, SSD has much better accuracy even with a smaller input image size. Code is available at https://github.com/weiliu89/caffe/tree/ssd.

/pdf/ssd-single-shot-multibox-detector-2251optntn.pdf

SSD: Single Shot MultiBox Detector

We present a class of efficient models called MobileNets for mobile and embedded vision applications. MobileNets are based on a streamlined architecture that uses depth-wise separable convolutions to build light weight deep neural networks. We introduce two simple global hyper-parameters that efficiently trade off between latency and accuracy. These hyper-parameters allow the model builder to choose the right sized model for their application based on the constraints of the problem. We present extensive experiments on resource and accuracy tradeoffs and show strong performance compared to other popular models on ImageNet classification. We then demonstrate the effectiveness of MobileNets across a wide range of applications and use cases including object detection, finegrain classification, face attributes and large scale geo-localization.

MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications

Scale variation has been a challenge from traditional to modern approaches in computer vision. Most solutions to scale issues have similar theme: a set of intuitive and manually designed policies that are generic and fixed (e.g. SIFT or feature pyramid). We argue that the scale policy should be learned from data. In this paper, we introduce ELASTIC, a simple, efficient and yet very effective approach to learn instance-specific scale policy from data. We formulate the scaling policy as a non-linear function inside the network’s structure that (a) is learned from data, (b) is instance specific, (c) does not add extra computation, and (d) can be applied on any network architecture. We applied ELASTIC to several state-of-the-art network architectures and showed consistent improvement without extra (sometimes even lower) computation on ImageNet classification, MSCOCO multi-label classification, and PASCAL VOC semantic segmentation. Our results show major improvement for images with scale challenges e.g. images with several small objects or objects with large scale variations. Our code and models will be publicly available soon.

ELASTIC: Improving CNNs with Instance Specific Scaling Policies.

With the recent progress in visual recognition, we have already started to see a surge of vision related real-world applications. These applications, unlike general scene understanding, are task oriented and require specific information from visual data. Considering the current growth in new sensory devices, feature designs, feature learning methods, and algorithms, the search in the space of features and models becomes combinatorial. In this paper, we propose a novel cost-sensitive task-oriented recognition method that is based on a combination of linguistic semantics and visual cues. Our task-oriented framework is able to generalize to unseen tasks for which there is no training data and outperforms state-of-the-art cost-based recognition baselines on our new task-based dataset.

/pdf/a-task-oriented-approach-for-cost-sensitive-recognition-4lxmn8l6tl.pdf

A Task-Oriented Approach for Cost-Sensitive Recognition

Deep down in many applications like natural language processing (NLP), vision, and robotics is a form of the k-nearest neighbor search algorithm (kNN). The kNN algorithm is primarily bottlenecked by data movement, limiting throughput and incurring latency in these applications. While there do exist well bounded kNN approximations that improve the performance of kNN, these algorithms trade-off accuracy and quickly degrade into linear search for high dimensionality. 
To address data movement, we designed the nearest neighbor content addressable memory (NCAM) which employs processing in-memory (PIM) to eliminate costly data transfers, and provides exact nearest neighbor search. NCAMs benefit from the modularity offered by 3D die-stacking technology which allows us to side-step the issues of direct integration with DRAM dies. The NCAM benefits from the higher density and speeds of emerging memory technologies and interfaces. 
We characterize a state-of-the-art software kNN implementation and expose the shortcomings of approximate kNN search. We present a full NCAM design and estimate its performance using post-placement and route estimates, and we show that its power characteristics are compatible with emerging memory substrates. We then evaluate energy efficiency and latency compared to modern multi-core CPUs and GPGPU platforms using parameters typical of mobile and server workloads. Our simulation results show that the NCAM can achieve up to 160x throughput per watt and three orders of magnitude latency improvement over a NVIDIA Titan X GPU for server workloads, and ~2413x throughput per watt and ~94.5x latency improvements over a multi-core Intel E5-2620 CPU. Finally, we show that the NCAM is not limited to just kNN and can be generalized to act as a content addressable memory (CAM) or ternary-CAM (TCAM).

NCAM: Near-Data Processing for Nearest Neighbor Search

Recent observations have advanced our understanding of the neural network optimization landscape, revealing the existence of (1) paths of high accuracy containing diverse solutions and (2) wider minima offering improved performance. Previous methods observing diverse paths require multiple training runs. In contrast we aim to leverage both property (1) and (2) with a single method and in a single training run. With a similar computational cost as training one model, we learn lines, curves, and simplexes of high-accuracy neural networks. These neural network subspaces contain diverse solutions that can be ensembled, approaching the ensemble performance of independently trained networks without the training cost. Moreover, using the subspace midpoint boosts accuracy, calibration, and robustness to label noise, outperforming Stochastic Weight Averaging.

/pdf/learning-neural-network-subspaces-1zieteqqjx.pdf

Learning Neural Network Subspaces

Enabling robust intelligence in the real-world entails systems that offer continuous inference while learning from varying amounts of data and supervision. The machine learning community has organically broken down this challenging goal into manageable sub-tasks such as supervised, few-shot, and continual learning. In light of substantial progress on each sub-task, we pose the question, "How well does this progress translate to more practical scenarios?" To investigate this question, we construct a new framework, FLUID, which removes certain assumptions made by current experimental setups while integrating these sub-tasks via the following design choices -- consuming sequential data, allowing for flexible training phases, being compute aware, and working in an open-world setting. Evaluating a broad set of methods on FLUID leads to new insights including strong evidence that methods are overfitting to their experimental setup. For example, we find that representative few-shot methods are substantially worse than simple baselines, self-supervised representations from MoCo fail to learn new classes when the downstream task contains a mix of new and old classes, and pretraining largely mitigates the problem of catastrophic forgetting. Finally, we propose two new simple methods which outperform all other evaluated methods which further questions our progress towards robust, real-world systems. Project page: this https URL.

Ali Farhadi

Papers

ELASTIC: Improving CNNs with Instance Specific Scaling Policies.

A Task-Oriented Approach for Cost-Sensitive Recognition

NCAM: Near-Data Processing for Nearest Neighbor Search

Learning Neural Network Subspaces

Are We Overfitting to Experimental Setups in Recognition