News
Views
Government
On campus
Noticed
In the harbour
Footnotes


News

1. Land titles

Screen Shot 2016-02-22 at 8.48.35 AM

Just days after the Nova Scotia Community College journalism class published its “untitled” investigation — which explores the historic injustice done to Black refugees who were promised property in Nova Scotia but were never given title to the land they settled, and how that injustice continues to impact the residents of North Preston — the province has responded with a promise to address the issue, reports the CBC:

“We understand that the community is concerned and we are in the process of finalizing a plan to address this matter, which has been around for a long, long time in the province,” Natural Resources Minister Lloyd Hines said Thursday.

He said the province is coming up with a process to “put it to bed,” and is partnering with Nova Scotia Legal Aid and the Nova Scotia Barristers’ Society, which is offering its services pro bono.

Asked about whether the province would help fund the process, Hines said there’s an “in-kind contribution to make” but wouldn’t say if money will be set aside. He said the plan will be completed “pretty immediately.”

Congratulations to the students. They put a lot of hard work into this project, shed light on an ongoing injustice, explained the problem in an engaging fashion, and as a result brought about real change. This is what journalism is supposed to do.

For this week’s Examineradio podcast, I interviewed Erin Moore, the class instructor, and Kristen Brown, one of the students involved in the project. The podcast will be published this afternoon.

2. Danger selfies

Steve Smith, putting a young person's life in danger, by taking a selfie.
Steve Smith, putting a young person’s life in danger by taking a selfie.

The Cape Breton and Central Nova Scotia Railway rail line between Port Hawkesbury and Sydney is probably history — the last train ran on the line since December 30, 2014, and Genesee & Wyoming, the company that owns the track, will apply to abandon the line in April.

But that hasn’t deterred people in Cape Breton: activists have formed the Scotia Rail Development Society in hopes of saving the line, and to raise awareness of their cause the group sponsored a “selfie contest“:

Participants are asked to create a selfie with abandoned tracks in the background with a message stating the significance of the rail and share it with hash-tag #SaveCBRail.

The contest ended last week, and a winner will be announced Sunday. Top prize is two movie tickets and a free dinner at Governor’s Pub.

All very simple, right? Well, no:

“The Scotia Rail Development Society board was surprised to receive a letter from Genesee & Wyoming forbidding young people from taking pictures within 30 feet of the track,” said society spokesman Greg MacLeod.

[…]

In its response, the company noted under transportation laws, it is illegal to trespass on railway property, under any circumstance, even if the track seems abandoned.

“Not only is it illegal, but it is also very dangerous,” said the company spokesperson.

There is exactly zero chance that anyone will be hit by a nonexistent train, but I suppose somebody might trip on the weeds growing near the tracks, so there’s that.

3. Airport

1

Stanfield International had 3,042,188 passengers last year, the airport said in a press release.

It would’ve been 3,042,231 but 133 potential passengers didn’t quite make it.

4.  “Mi’kmaq Cinderella”

Photo: Nova Scotia Museum/novascotia.ca
Photo: Nova Scotia Museum/novascotia.ca

The above rock carving is found in Kejimkujik National Park. It’s the subject of an interesting article on the CBC site today.

5. Visitor information centres

The province will keep the six remaining visitor information centres open, Department of Business minister Mark Furey announced yesterday.


Views

1. Mulgrave Park

1

Peter Ziobrowski explains how Mulgrave Park was conceived and built.

2. Banshees

Daniel MacDonald documents the convention centre banshees.

3. Cranky letter of the day

To the New Glasgow News:

I was out for a coffee at Tim’s and read a comment from what sounded like ‘The Economic Disaster,’ so I made my own little version.

Here’s another quiz, and if you answer correctly you can continue to live happily in the County.

1 – What will enable us to survive and thrive after five years when the amalgamation funding runs out, equalization payments change and the ‘One Pictou County Municipality’ debts pile up?

2 – Aren’t the County finances in better shape than some others?

3 – Doesn’t the Municipality of Pictou County continue to prioritize spending in small communities as necessary?

4 – What fictional reality will allow the unification without increasing taxation?

5 – Do we continue to have DOT depots here that plough our County roads quicker than most?

6 – Do we want to amalgamate and practically advertise ‘Free Money’ to attract business?    What is ‘New Scotland Business Development Inc. anyway?

7 – Do we have great volunteer fire departments that improve and develop new co-operative initiatives?

8 – Do we have local councillors who provide personalized service in an efficient and effective way?

9 – Does bigger necessarily mean better?

10 – Isn’t the County a better place to live?

If it ain’t broke don’t fix it!

Vote No on Amalgamation

Ken Fraser, RR2 Hopewell


Government

No public meetings.


On Campus

Dalhousie

Thesis defence, Political Science (10:30am, Room 3107, Mona Campbell Building) — Glenn Graham will defend his thesis, “Regionalism on the Celtic Fringe: How a Peripheral Community Resists, Negotiates, and Accommodates Political and Economic Integration.”

Semantic image representation (11:30am, Slonim Conference Room, Goldberg Computer Science Building) — PhD student Elham Etemad will present on “Semantic image representation for visual classification based on N-grams of visual words with shape and spatial pyramids.” No, I don’t know what that means either. Here’s what she has to say:

As the rapid evolution of advanced and affordable visual sensors, which are often embedded in various powerful mobile devices, people nowadays can easily capture images or videos, or access massive collections of visual data online. This creates huge application opportunities, from sensor based information retrieval to harvesting meaningful information and knowledge from large image or video datasets via machine learning. A typical computation task for such applications is image classification, in that an image can be automatically classified and recognized against a large number of prior tags or textural class descriptions, either generic or application specific, via machine learning or manual annotation process. Many image representation methods were introduced to tackle this task, commonly categorized into Local Image Descriptors, Global Image Descriptors, and Learned Network Representations. While some of these methods achieve promising results, they still ! suffer from the Semantic Gap, i.e. lacking higher level semantics within the representation due to relying on low level features and neglecting semantic properties, or coded in learned network without transparency. As a result, it is difficult to apply such representation to support visual semantic based reasoning for knowledge based systems. In addition, the existing representation methods often require to compute a large number of parameters for classification, which may add a prohibitive cost to a real-time image or video analysis and reasoning system.

In this research, we introduce a new image representation method, based on N-games of Perceptual Shape Tokens (PST), organized in Shape-Spatial Pyramids for coding both local and global shape semantics of images. The PST can be used to form multiple layers of semantics of image representation for minimizing the semantic gap between low level image representation and its class description. PST are the baseline visual words for creating higher level semantic visual descriptors. A Shape Pyramid consists of multi-levels of visual words obtained based on the N-gram notion of PST. It also adds semantics to the image representation by considering the syntactic relationships between N-gram PST words. Considering that most of the current BoVW based methods do not apply the spatial layout and regularities of visual words in images, we integrate the Spatial Pyramid with the Shape Pyramid for the shape-spatial distribution information of image semantics. The proposed method produces improved classification results on Wang and Caltech datasets compared with the state-of-the-art methods.

Quakerism (3:30pm, Marion McCain Building, room 1170) — Brennan Dempsey will speak on “‘Here I Die Innocently’: Children and Youth in early Quakerism.”

Multiple Sclerosis (3:40pm, LSC 5260) — Virender Bhan will speak on “Multiple Sclerosis: then and now.”


Noticed

Have a Tarantula Nebula:

1


In the harbour

The seas around Nova Scotia, 8:20am Friday. Map: marinetraffic.com
The seas around Nova Scotia, 8:20am Friday. Map: marinetraffic.com

Dinkeldiep, cargo, Saint-Pierre to Pier 41

Atlantic Star sails to sea
Fritz Reuter sails to sea


Footnotes

TGIF.

Tim Bousquet is the editor and publisher of the Halifax Examiner. Twitter @Tim_Bousquet Mastodon

Join the Conversation

10 Comments

Only subscribers to the Halifax Examiner may comment on articles. We moderate all comments. Be respectful; whenever possible, provide links to credible documentary evidence to back up your factual claims. Please read our Commenting Policy.
  1. Liberals are in election season. Trying to win the votes back. Look for sucking up the next 2 years.
    Liberals also want to privatize everything. NSLC just announced their big profits this week. Liberals will want to make sure that money goes into private hands pretty soon – the government doesn’t need it, they can make more cuts to cover it so we will all be rich.

    1. Food for thought: Something like 75% of alcohol is consumed by the top 10% of drinkers. I’m not pro-prohibition, but profiting off of alcoholics is a morally and economically (if you have to pay for the consequences of all that drinking) questionable practice.

    2. Yeah Gina, just think if government operated grocery stores, restaurants, airlines, clothes stores etc the province would be full of happy prosperous people.
      If it makes you feel better I prefer the NSLC be sold to the civil service pension fund.

  2. Thanks for publishing minister Furey’s announcement about the visitor centres. Perhaps the govt info folks would be considered less of a “bullshit machine” if they would be more forthcoming in their announcements. At the post-cabinet media scrum immediately following the news release minister Furey admitted that the govt “needs” to privatize those centres. Would have been more “honest” of govt to include that info – which they clearly knew – in the news release. http://www.southcoasttoday.ca/content/visitor-centres-need-be-privatized-says-minister

  3. The rail line properties and infrastructure should be retained and not sold off to some developer or turned into some kind of “permanent” active transportation or recreational trail network. Halifax is a good example why this is so; today when Halifax is continually trying to find ways to establish more effective public transportation routes, they must wonder about the logic of re-purposing the rail lines that could have filled a role in providing a faster public transit service. Cape Breton may not be able to justify operating the rail line today, but the right of ways (RoW) that exist will be next to impossible to recreate in the future. Do not allow the rail RoW to be eliminated in order to recoup a minor monetary gain by selling the land off. The rail RoW is a very valuable asset that cannot be easily recreated nor can it afford to be lost.

  4. Glad somebody posted the convention centre banshees to the internets- when I was waiting for the crosswalk at Brunswick and Sackville, all of us standing were really captivated by it! Very haunting.

  5. The semantic image talk appears to be on computer vision and an approach to get better results by letting the algorithm determine context around what its reading

    902 is more likely than 9O2 for example bit both are valid solutions.

    1. My take on it is this: Current image detection routines work by attempting to tell you what kind of book you are reading by analyzing all the words in the book, but the words currently can’t be read in order. This algorithm improves upon that by collecting the words into sentences – which are still not in order (the sentences themselves are not in order, and likewise, the words in the sentences are in random order), but it’s a big improvement!

    2. I initially made an effort to write my Ph.D. thesis using relatively accessible language and was accused of trying to “insult” my advisor as a result. Because no one besides my thesis defence committee was ever going to read the damn thing and they didn’t need the big words explained to them.