“Users want control” is a shoulder shrug

“Users want control” is a shoulder shrug

So when we say users want control over X – their privacy, their security, their data, their history – we are first acknowledging that current systems act against users, but we aren’t proposing any real solution. For instance, we say “users want control over their privacy,” but what people really want is some subset of:

There’s no easy answers, not everyone holds all these desires, but these are concrete ways of thinking about what people want. When we simply say users want control it’s giving up on understanding people’s specific desires.

Source: www.ianbicking.org

Baltimore city government computer network hit by ransomware attack

Baltimore city government computer network hit by ransomware attack

Baltimore city government computers were infected with ransomware Tuesday, the mayor’s office said, the second time in just over a year that hackers demanding payment disrupted the city’s technology systems. More: Ransomware attack on Baltimore government computers ‘very serious issue’ »

Davis said critical systems, including 911 and 311, were not been affected, but that the majority of city servers were shut down. The Baltimore Sun obtained a copy of a ransom note left on a Baltimore city computer; it also identified the ransomware as RobbinHood.

Source: www.baltimoresun.com

Deadly Japanese earthquake study retracted over false data

Deadly Japanese earthquake study retracted over false data

Authors of a study in Science analysing an earthquake that struck Kumamoto in southern Japan in 2016 have retracted their paper, following a university investigation that found it contained falsified data and manipulated images. The paper, published in November 2016, purported to show that Mount Aso, one of the world’s largest active volcanoes and situated roughly 30 kilometres northeast of Kumamoto City, had halted the rupture zone of a magnitude-7.0 earthquake on 16 April, and prevented it from spreading further and causing more damage than it did. The paper’s data and figures showed a 40-kilometre-long rupture zone that ended inside Mount Aso’s caldera, a depression that forms when a volcano collapses in on itself, and at a depth consistent with the volcano’s magma chamber.

Source: www.nature.com

Human Pose Estimation with Deep Learning

Human Pose Estimation with Deep Learning

Images are cropped around the predicted joint and fed to the next stage, in this way the subsequent pose regressors see higher resolution images and thus learn features for finer scales which ultimately leads to higher precision. The model is trained by minimizing the Mean Squared-Error (MSE) distance of our predicted heat-map to a target heat-map (The target is a 2D Gaussian of constant variance (σ ≈ 1.5 pixels) centered at the ground-truth \((x, y)\) joint location)

g1() and g2() predict heatmaps (belief maps in the paper). The HRNet (High-Resolution Network) model has outperformed all existing methods on Keypoint Detection, Multi-Person Pose Estimation and Pose Estimation tasks in the COCO dataset and is the most recent.

Source: blog.nanonets.com

Show HN: 2D Graphics on Modern GPU

Show HN: 2D Graphics on Modern GPU

The most promising use case for the work is likely CAD tools, where there might be quite complex scenes and it’s not necessarily practical to organize the UI drawing around GPU primitives (as opposed to 3D games, for example). Similar to the simplistic rendering strategy above, a simple approach to tile generation would be to have a thread per tile (~12k threads), each of which sequentially traverses the scene graph. The basic structure is there, though; the traversal of the scene graph and generation of tiles is at heart sequential, not relying on tricky GPU-compute techniques such as sorting.

Source: raphlinus.github.io

A Nazi Critic and a Gestapo Spy

A Nazi Critic and a Gestapo Spy

In September 1944, the Nazis murdered the teacher Elisabeth von Thadden, the aunt of DER SPIEGEL editor Marianne Wellershoff. DER SPIEGEL: Ms. Ruppel, you met Elisabeth von Thadden for the first time in 1937 when you started going to her Protestant boarding school in Wieblingen, a district of the southern German city of Heidelberg. DER SPIEGEL: Did you maintain contact with Elisabeth von Thadden following your year at the Wiebling boarding school?

Source: www.spiegel.de

Svgbob: Convert your ASCII diagram scribbles into happy little SVG

Svgbob: Convert your ASCII diagram scribbles into happy little SVG

Svgbob is an ascii to svg converter. If you like this project please consider supporting me on patreon

normally operates on stdin and stdout:

produces an SVG in similar to the one produced by the old binary. also allows passing arguments instead:

And you can mix and match:

You can pipe string from the output of other program:

You can also pass an inline string:

This is also documented in the output of :

SvgBobRus is an ascii to svg converter –font-family text will be rendered with this font (default: ‘arial’) –font-size text will be rendered with this font size (default: 14) –scale scale the entire svg (dimensions, font size, stroke width) by this factor help Prints this message or the help of the given subcommand(s)

The SVG image is generated from the ascii text below.

Source: github.com

A 2019 Guide to Human Pose Estimation with Deep Learning

A 2019 Guide to Human Pose Estimation with Deep Learning

Images are cropped around the predicted joint and fed to the next stage, in this way the subsequent pose regressors see higher resolution images and thus learn features for finer scales which ultimately leads to higher precision. The model is trained by minimizing the Mean Squared-Error (MSE) distance of our predicted heat-map to a target heat-map (The target is a 2D Gaussian of constant variance (σ ≈ 1.5 pixels) centered at the ground-truth \((x, y)\) joint location)

g1() and g2() predict heatmaps (belief maps in the paper). The HRNet (High-Resolution Network) model has outperformed all existing methods on Keypoint Detection, Multi-Person Pose Estimation and Pose Estimation tasks in the COCO dataset and is the most recent.

Source: blog.nanonets.com

Driven – A bicycle drivechain concept without derailleurs and chains

Driven – A bicycle drivechain concept without derailleurs and chains

To achieve ultimate efficiency, Driven utilises 21 CeramicSpeed hybrid bearings. Each bearing contains unique ceramic balls that are 58% lighter and 2.3 times harder than traditional steel, increasing the speed of the bearing by 30-50%. Each bearing works in harmony to deliver greater power, playing a pivotal role in performance and speed.

Source: www.ceramicspeed.com