Most existing crowd counting methods require object location-level annotation, i.e., placing a dot at the center of an object. While being simpler than the bounding-box or pixel-level annotation, obtaining this annotation is still labor-intensive and time-consuming especially for images with highly crowded scenes. On the other hand, weaker annotations that only know the total count of objects can be almost effortless in many practical scenarios. Thus, it is desirable to develop a learning method that can effectively train models from count-level annotations. To this end, this paper studies the problem of weakly-supervised crowd counting which learns a model from only a small amount of location-level annotations (fully-supervised) but a large amount of count-level annotations (weakly-supervised). To perform effective training in this scenario, we observe that the direct solution of regressing the integral of density map to the object count is not sufficient and it is beneficial to introduce stronger regularizations on the predicted density map of weakly-annotated images. We devise a simple-yet-effective training strategy, namely Multiple Auxiliary Tasks Training (MATT), to construct regularizes for restricting the freedom of the generated density maps. Through extensive experiments on existing datasets and a newly proposed dataset, we validate the effectiveness of the proposed weakly-supervised method and demonstrate its superior performance over existing solutions.

Batch normalization (BatchNorm) has become an indispensable tool for training deep neural networks, yet it is still poorly understood. Although previous work has typically focused on its normalization component, BatchNorm also adds two per-feature trainable parameters: a coefficient and a bias. However, the role and expressive power of these parameters remains unclear. To study this question, we investigate the performance achieved when training only these parameters and freezing all others at their random initializations. We find that doing so leads to surprisingly high performance. For example, a sufficiently deep ResNet reaches 83% accuracy on CIFAR-10 in this configuration. Interestingly, BatchNorm achieves this performance in part by naturally learning to disable around a third of the random features without any changes to the training objective. Not only do these results highlight the under-appreciated role of the affine parameters in BatchNorm, but - in a broader sense - they characterize the expressive power of neural networks constructed simply by shifting and rescaling random features.

Running deep neural network (DNN) inference on mobile devices, i.e., mobile inference, has become a growing trend, making inference less dependent on network connections and keeping private data locally. The prior studies on optimizing DNNs for mobile inference typically focus on the metric of average inference latency, thus implicitly assuming that mobile inference exhibits little latency variability. In this note, we conduct a preliminary measurement study on the latency variability of DNNs for mobile inference. We show that the inference latency variability can become quite significant in the presence of CPU resource contention. More interestingly, unlike the common belief that the relative performance superiority of DNNs on one device can carry over to another device and/or another level of resource contention, we highlight that a DNN model with a better latency performance than another model can become outperformed by the other model when resource contention be more severe or running on another device. Thus, when optimizing DNN models for mobile inference, only measuring the average latency may not be adequate; instead, latency variability under various conditions should be accounted for, including but not limited to different devices and different levels of CPU resource contention considered in this note.

Reinforcement learning has shown much success in games such as chess, backgammon and Go [1, 2, 3]. However, in most of these games, agents have full knowledge of the environment at all times. In this paper, we describe a deep learning model that successfully optimizes its score using reinforcement learning in a game with incomplete and imperfect information. We apply our model to FlipIt [4], a two-player game in which both players, the attacker and the defender, compete for ownership of a shared resource and only receive information on the current state (such as the current owner of the resource, or the time since the opponent last moved, etc.) upon making a move. Our model is a deep neural network combined with Q-learning and is trained to maximize the defender's time of ownership of the resource. Despite the imperfect observations, our model successfully learns an optimal cost-effective counter-strategy and shows the advantages of the use of deep reinforcement learning in game theoretic scenarios. Our results show that it outperforms the Greedy strategy against distributions such as periodic and exponential distributions without any prior knowledge of the opponent's strategy, and we generalize the model to n-player games.

We propose a method of training quantization thresholds (TQT) for uniform symmetric quantizers using standard backpropagation and gradient descent. Contrary to prior work, we show that a careful analysis of the straight-through estimator for threshold gradients allows for a natural range-precision trade-off leading to better optima. Our quantizers are constrained to use power-of-2 scale-factors and per-tensor scaling of weights and activations to make it amenable for hardware implementations. We present analytical support for the general robustness of our methods and empirically validate them on various CNNs for ImageNet classification. We are able to achieve near-floating-point accuracy on traditionally difficult networks such as MobileNets with less than 5 epochs of quantized (8-bit) retraining. Finally, we present Graffitist, a framework that enables automatic quantization of TensorFlow graphs for TQT (available at https://github.com/Xilinx/graffitist ).

Deep Learning Market report to study and analyses the market size (Consumption, Value, Volume and Production) By Company, Key Regions, Products and End User/Application, Deep Learning market breakdown data from 2014 to 2019, and 6 year forecast from 2020 to 2026. Bedsides Deep Learning industry research report enriched on worldwide competition by topmost prime manufactures (Amazon Web Services (AWS), Google, IBM, Intel, Micron Technology, Microsoft, Nvidia, Qualcomm, Samsung Electronics, Sensory Inc., Skymind, Xilinx, AMD, General Vision, Graphcore, Mellanox Technologies, Huawei Technologies, Fujitsu, Baidu, Mythic, Adapteva, Inc., Koniku) which providing information such as Company Profiles, Gross, Gross Margin, Capacity, Product Picture and Specification, Production, Price, Cost, Revenue and contact information.Deep Learning Market report provide the in-depth analysis of key factors influencing the growth of the market (Growth Potential, Opportunities, Drivers, Industry-Specific Challenges and Risks). The Latest Deep Learning Industry Data Included in this Report: Deep Learning Market Size & Analysis (2014 – 2026); Deep Learning Market Volume & Future Trends (2014 – 2026); Deep Learning Market; By Geography (Volume and Value); 2014 – 2026; Deep Learning Market Opportunity Assessment (2014 – 2026); Deep Learning (Installed Base) Market Share: By Company; Major Deals in Deep Learning Market; Deep Learning Reimbursement Scenario; Deep Learning Current Applications; Deep Learning Competitive Analysis: By Company; Key Market Drivers and Inhibitors; Major Companies Analysis. Scope of Deep Learning Market: The deep learning market has been segmented on the basis of offerings, applications, end-user industries, and geographies. In terms of offerings, software holds the largest share of the deep learning market.

We introduce the weightwatcher (ww), a python tool for a python tool for computing quality metrics of trained, and pretrained, Deep Neural Netwworks. This blog describes how to use the tool in practice; see our most recent paper for even more details. The summary contains the Power Law exponent (), as well as several log norm metrics, as explained in our papers, and below. Each value represents an empirical quality metric that can be used to gauge the gross effectiveness of the model, as compared to similar models. We can use these metrics to compare models across a common architecture series, such as the VGG series, the ResNet series, etc. These can be applied to trained models, pretrained models, and/or even fine-tuned models.

Your code is more complicated than you think. One of the first things every software developer learns about is the command-line. At its core, the command-line is a list of strings that are typically broken down into flags (e.g., -- verbose) and arguments (e.g., -- port 80). This is enough for many simple applications. You can define 2 to 3 command-line arguments in a command-line interface (CLI) parsing library, and you are done.

We describe libmolgrid, a general-purpose library for representing three-dimensional molecules using multidimensional arrays of voxelized molecular data. It was designed for seamless integration with popular deep learning frameworks and features optimized performance by leveraging graphics processing units (GPUs).

"We are transforming the business analysis process through automation, aimed at making the life of every analyst focused on providing quality output rather than substantial manual efforts on data quality check, " Ramesh Lakshminarayanan, CIO, CRISIL told ETCIO. To improve its research reports and analytics, Crisil has been adopting automated data extraction including extraction of unstructured paragraphs, tables, etc which are automated to the data mapping (financial taxonomy mapping) and automatic text summarisation (Natural Language Generation, or NLG) of analyst opinion. Almost 90 percent of Crisil's key processes are data-driven. The company has initiated a large streamlining effort in 2019, in multiple phases. The first phase was to automate a lot of mundane web crawling/ extraction work, as the ratings agency collected a chunk of data from various sources on the web.