Two series of single and stereo photographs display a range of natural conditions and fuel loadings in post-hurricane forests in the southeastern United States. Each group of photos includes inventory information summarizing vegetation composition, structure and loading, woody material loading and density by size class, forest floor loading, and various site characteristics. The natural fuels photo series is designed to help land managers appraise fuel and vegetation conditions in natural settings.

Forest managers are seeking practical guidance on how to adapt their current practices and, if necessary, their management goals, in response to climate change. Science-management collaboration was initiated on national forests in eastern Washington where resource managers showed a keen interest in science-based options for adapting to climate change at a 2-day workshop. Scientists and managers reviewed current climate change science and identified resources vulnerable to expected climate change. Vulnerabilities related to vegetation and habitat management included potential reductions in forest biodiversity and low forest resilience to changing disturbance regimes. The vulnerabilities related to aquatic and infrastructure resources included changing water quality and quantity, the risk to roads and other facilities from changes to hydrologic regimes, and the potential loss of at-risk aquatic species and habitats. Managers then worked in facilitated groups to identify adaptations that could be implemented through management and planning to reduce the vulnerability of key resources to climate change. The identified adaptations were grouped under two major headings: Increasing Ecological Resiliency to Climate Change, and Increasing Social and Economic Resiliency to Climate Change. The information generated from the science-management collaborative represents an initial and important step in identifying and prioritizing tangible steps to address climate change in forest management. Next would be the development of detailed implementation strategies that address the identified management adaptations..

Integrated research is about achieving holistic understanding of complex biophysical and social issues and problems. It is driven by the need to improve understanding about such systems and to improve resource management by using the results of integrated research processes. Traditional research tends to fragment complex problems, focusing more on the pieces of problems rather than the whole that comprises multiple interrelationships and interactions. The outcome is that a lot is known about the parts (e.g., recreation, fish, and wildlife) but relatively little about how they are interrelated. There seems to be general agreement that integrated questions must drive the search for integrated understanding, but tradition, inertia, institutional culture,budgets, training, and lack of effective leadership foster reductionism (at worst) or minimal degrees of integration (at best) rather than any substantial, sustainable effort toward integrated research. In this paper, a phased approach to framing integrated research questions and addressing the substantial barriers that impede integrated efforts are discussed. A key conclusion is that to make any significant progress toward comprehensive integrated research will require more than rhetoric. Progress must begin with more effective leadership throughout various levels of research organizations.

This guidebook describes the High Peak/Moon Creek Research Natural Area, a 617.5-ha (1,526-ac) tract of coniferous forest containing stands dominated by 100- to 150-year-old Douglas-fir, a small old-growth (500+ years) Douglas-fir stand, and riparian vegetation within the western hemlock zone of the Coast Range in western Oregon.

Roger Lake Research Natural Area (RNA), a 174.7-ha reserve in north-central Washington, contains a rich diversity of landforms, plant communities, and wildlife habitats. Spreading outward from the lake itself, sedge and sphagnum fens give way to upland coniferous forest, granitic cliffs, and a relictual, high-altitude big sagebrush-whitebark pine (Artemisia tridentata-Pinus albicaulis) meadow. Five sensitive plant species and several vertebrate species that are rare in the region occur in the RNA. Dynamic ecological processes in action in the RNA are revealed in the paludification of the forest edge; aging, broken beaver dams; and widespread bark beetle-induced conifer mortality.

National forest lands encompass numerous rural and urban communities. Some national-forest-based communities lie embedded within national forests, and others reside just outside the official boundaries of national forests. The urban and rural communities within or near national forest lands include a wide variety of historical traditions and cultural values that affect their process of economic development. National-forest-based urban and rural communities participate in numerous economic sectors including nontraded industries, resource-dependent traded industries, and non-resource-dependent traded industries. These communities represent microeconomic environments. Cluster theory provides an explicit framework to examine the microeconomic relationships between national forests and their embedded and neighboring communities. Implementation of economic cluster initiatives in national-forest-based communities could improve their overall social well-being through increased competitive advantage based on innovation and higher productivity. This paper proposes establishing an Economic Clusters research team within the Forest Service. This team would dedicate its efforts to the analysis and improvement of the determinants of competitive advantage affecting national-forest-based communities.

A series of single and stereo photographs display a range of natural conditions and fuel loadings in evergreen and deciduous oak/juniper woodland and savannah ecosystems in southern Arizona and New Mexico. This group of photos includes inventory data summarizing vegetation composition, structure, and loading; woody material loading and density by size class; forest floor coverage and loading; and various site characteristics. The natural fuels photo series is designed to help land managers appraise fuel and vegetation conditions in natural settings.

Two series of single and stereo photographs display a range of natural conditions and fuel loadings in sagebrush with grass and ponderosa pinejuniper types in central Montana. Each group of photos includes inventory information summarizing vegetation composition, structure, and loading; woody material loading and density by size class; forest floor depth and loading; and various site characteristics. The natural fuels photo series is designed to help land managers appraise fuel and vegetation conditions in natural settings.

Silvicultural practices in the Douglas-fir region evolved through a combination of formal research, observation, and practical experience of forest managers and silviculturists, and changing economic and social factors. This process began more than a century ago and still continues. It has had a great influence on the economic well-being of the region and on the present characteristics of the region's forests. This long history is unknown to most of the public, and much of it is unfamiliar to many natural resource specialists outside (and even within) the field of silviculture. We trace the history of how we got where we are today and the contribution of silvicultural research to the evolution of forest practices. We give special attention to the large body of information developed in the first half of the past century that is becoming increasingly unfamiliar to both operational foresters and--perhaps more importantly--to those engaged in forestry research. We also discuss some current trends in silviculture and silviculture-related research.

This guidebook describes the Forest Peak Research Natural Area (RNA), a 62.8-ha (153.3-ac) tract containing a mature Douglas-fir (Pseudotsuga menziesii) forest and a grass bald within the Willamette Valley Foothill Ecoregion. Forest Peak RNA also contains an undisturbed third-order stream reach.

This guidebook describes the Saddle Bag Mountain Research Natural Area, a 121-ha (300-ac) tract established to represent an old-growth remnant of Pacific silver fir (Abies amabilis) and western hemlock (Tsuga heterophylla) forest in the Oregon Coast Range. Pacific silver fir and noble fir (Abies procera) occur as isolated remnants, and both species are approaching the southern limits of their natural range in the Oregon Coast Range.

This guidebook describes the Grass Mountain Research Natural Area, a 377-ha (931-ac) tract in the Oregon Coast Range. The area supports a grass bald complex surrounded by stands dominated by noble fir (Abies procera) and/or Douglas-fir (Pseudotsuga menziesii) in the overstory, and western hemlock (Tsuga heterophylla) in the understory. The area also contains a small rock garden plant community along high-elevation ridges, and young Douglas-fir forest that originated from a wildfire. Headwaters of high-elevation, Oregon Coast Range streams are surrounded by noble fir forest and add to the site diversity.

This report examines the growth and development of the tourism industry in Hoonah, Alaska, and its effects on community life and resource use. The report describes the gradual development of tourism in Hoonah and presents resident perceptions of tourism's effect on the natural and social environment. A multisited ethnographic approach was used featuring indepth, open-ended interviews with local residents, tourism providers, business owners, and government officials. Data were analyzed using Ethnograph, a software program used to assist in coding data based on prominent themes. Results indicate that tourism has brought changes to the lives of Hoonah residents, particularly those relying on natural resources for everyday survival. Expansion of nature-based tourism in the area surrounding Hoonah resulted in conflicts between resource users. The growth of the charter fishing fleet led to competition with commercial fishers. Nature tour operators using remote recreation sites experienced conflicts with local hunters and fishers as well as other commercial guides. The development of a cruise ship destination on private land outside of Hoonah led to shifts in use of this historic site by local residents as well as in use of other private lands used for subsistence. Findings may enable Forest Service planners to identify factors involved in the relation between tourism growth and community well-being. It also may assist small southeast Alaska communities in decisionmaking related to tourism development.

This guidebook describes the Little Sink Research Natural Area, a 32.38-ha (80-ac) tract occupying an area of geologically unstable marine siltstone exhibiting natural geomorphic disturbances including landslides, slump benches, scarps, basins and ponds. The area supports forested stands dominated by Douglas-fir (Pseudotsuga menziesii) as well as stands codominated by Douglas-fir and bigleaf maple (Acer macrophyllum) representative of coniferous forest along the foothills of the Willamette Valley.

This report synthesizes the literature on the role of informal economic activity in the United States postindustrial economy. Informal economic activity is expanding in the United States and is likely to continue in the foreseeable future. The formal and informal economic sectors are inextricably intertwined, with individuals and households combining elements of both sectors to construct their livelihoods. Although the informal economy is often thought of as the domain of economically marginal individuals and households, virtually everyone participates in the informal economy to some extent. However, the literature highlights how factors such as social status and household position in the formal economy affect whether participation in informal economic activity is exploitative or empowering. The nontimber forest products sector serves as a case study of why it is important to consider informal economic activity when developing natural resource and economic development policy. We recommend steps policymakers can take to identify and encourage positive aspects of the informal economic activity. We also highlight several areas of research to improve understandings of the role of informal economic activity in postindustrial societies.

This paper describes a growth model for young plantations of Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) growing in the Pacific Northwest. The overall model has three major components. The first is a yield model for diameter and height distributions describing stands prior to pruning or precommercial thinning. The second component is an annual per-acre net increment model adapted from a recent model for Douglas-fir plantations in New Zealand; thinning and pruning are features of the model. The third component is growth equations for cohorts of individual trees; the results from this component are adjusted to match those from the second component. Fitting data are from Stand Management Cooperative experiments, with top heights generally below 75 ft. An intended use of the model is the evaluation of pruning regimes, in conjunction with the ORGANON model for growth at older ages, and TREEVAL model for clear-wood recovery and economic evaluation.

This guidebook describes Benjamin Research Natural Area, a 258-ha (637-ac) tract originally established to represent an example of the western juniper/Idaho fescue (Juniperus occidentalis/Festuca idahoensis) plant association. Subsequent field surveys indicate the predominant vegetation is best characterized as the western juniper/low sagebrush/Idaho fescue plant association. Current vegetation is dominated by western juniper woodland with an understory vegetation mosaic that varies with soil depth. Low sagebrush (Artemisia arbuscula Nutt.) occurs as the major shrub in shallow or rocky soils, and Wyoming big sagebrush (Artemisia tridentata Nutt. ssp. wyomingensis Beetle and Young) predominates in areas with deeper or more finely textured soil.

This guidebook describes Woodcock Bog Research Natural Area (RNA), a 114-ha (281-ac) area located within the Klamath-Siskiyou ecoregion in southwestern Oregon. The RNA includes a hanging fen and stream segment on ultramafic rock and derived soils. Numerous plant species occur within the fens that are endemic to the Klamath-Siskiyou Mountains of southwestern Oregon and northwestern California. Cobra lily (Darlingtonia californica), and sedges (Carex spp.) characterize the area. The site also supports very dry, open serpentine forest stands of Jeffrey pine (Pinus jeffreyi), along with denser stands of Port-Orfordcedar (Chamaecyparis lawsoniana), Douglas-fir (Pseudotsuga menziesii), and other conifers typical of the region.

This guidebook describes Cherry Creek Research Natural Area, a 239-ha (590-ac) area that supports old-growth Douglas-fir-western hemlock (Pseudotsuga menziesii- Tsuga heterophylla) forest occurring on sedimentary materials in the southern Oregon Coast Range.

This report is a compilation of four briefing papers based on literature reviews and syntheses, prepared for USDA Forest Service policy analysts and decisionmakers about specific questions pertaining to climate change. The main topics addressed here are effects of climate change on wildlife habitat, other ecosystem services, and land values; socioeconomic impacts of climate change on rural communities; and competitiveness of carbon offset projects on nonindustrial private forests in the United States.

This guidebook describes Camas Swale Research Natural Area, a 127-ha (314-ac) area that supports dry site, old-growth Douglas-fir (Pseudotsuga menziesii) forest. Major plant associations present within the area include the Douglas-fir/salal/western swordfern (Pseudotsuga menziesii/Gaultheria shallon/Polystichummunitum) plant association, Douglas-fir/Oregongrape (Pseudotsuga menziesii/Berberis nervosa) plant association, Douglas-fir/poison oak (Pseudotsuga menziesii/Toxicodendron diversilobum) plant association, and Douglas-fir/hazelnut-trailing snowberry/western swordfern (Pseudotsuga menziesii/Corylus cornutavar. californica-Symphoricarpos mollis/Polystichum munitum) plant association.

A gradient-analysis-based model and grid-based map are presented that use the potential vegetation zone as the object of the model. Several new variables are presented that describe the environmental gradients of the landscape at different scales. Boundary algorithms are conceptualized, and then defined, that describe the environmental boundaries between vegetation zones on the Olympic Peninsula, Washington, USA.

This guidebook describes Fox Hollow Research Natural Area (RNA), a 66-ha (163-ac) area that supports dry-site Douglas-fir (Pseudotsuga menziesii)–ponderosa pine (Pinus ponderosa) forest within the Oregon Coast Range ecoregion. Major forest plant associations represented at Fox Hollow RNA include Douglas-fir/salal/western swordfern (Pseudotsuga menziesii/Gaultheria shallon/Polystichum munitum) forest and Douglas-fir/Oregongrape (Pseudotsuga menziesii/Berberis nervosa) forest. Other forested communities are represented within the RNA in minor amounts including: Douglas-fir/poison oak (Pseudotsuga menziesii/ Toxicodendron diversilobum) forest, ponderosa pine-Douglas-fir/California fescue (Pinus ponderosa-Pseudotsuga menziesii/Festuca californica) woodland, and ponderosa pine-Douglas-fir-California black oak (Pinus ponderosa-Pseudotsuga menziesii-Quercus kelloggii) woodland.

Three series of photographs display a range of natural conditions and fuel loadings for sagebrush-steppe types that are ecotonal with grasses, western juniper, and ponderosa pine in eastern Oregon, and one series of photographs displays a range of natural conditions and fuel loadings for northern spotted owl nesting habitat in forest types in Washington and Oregon. Each group of photos includes inventory information summarizing vegetation composition, structure, and loading; woody material loading and density by size class; forest floor depth and loading; and various site characteristics. The natural fuels photo series is designed to help land managers appraise fuel and vegetation conditions in natural settings.

This guidebook describes Upper Elk Meadows Research Natural Area (RNA), a 90-ha (223-ac) area that supports a mixture of coniferous forest and open, shruband herb-dominated wetlands. The major forest plant association present within Upper Elk Meadows RNA is Pacific silver fir/vine maple/coolwort foamflower (Abies amabilis/Acer circinatum-Tiarella trifoliata).

We monitored coarse woody debris dynamics and natural tree regeneration over a 14-year period after the 1991 Warner Creek Fire, a 3631-ha (8,972-ac) mixed severity fire in the western Cascade Range of Oregon. Rates for tree mortality in the fire, postfire mortality, snag fall, and snag fragmentation all showed distinct patterns by tree diameter and species, with Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) more likely to survive a fire, and to remain standing as a snag, than other common tree species. Natural seedling regeneration was abundant, rapid, and highly variable in space. Densities of seedlings >10 cm height at 14 years postfire ranged from 1,530 to 392,000 per ha. Seedling establishment was not concentrated in a single year, and did not appear to be limited by the abundant growth of shrubs. The simultaneous processes of mortality, snag fall, and tree regeneration increased the variety of many measures of forest structure. The singular event of the fire has increased the structural diversity of the landscape.

These are the proceedings of the 2005 workshop titled implications of bringing climate into natural resource management in the Western United States. This workshop was an attempt to further the dialogue among scientists, land managers, landowners, interested stakeholders and the public about how individuals are addressing climate change in natural resource management.

This guidebook describes major biological and physical attributes of the 243-ha(600-ac) North Fork Silver Creek Research Natural Area (RNA), Josephine County, Oregon.

This guidebook describes major biological and physical attributes of the 243-ha (600-ac) Round Top Butte Research Natural Area. The area supports high-quality examples of valley upland grasslands and savanna of the Cascade foothills. Plant communities include Oregon white oak (Quercus garryana) savanna and open woodland with forbs and grasses; ponderosa pine (Pinus ponderosa)-Oregon white oak woodland; bluebunch wheatgrass-California oatgrass-Lemmon’s needlegrass (Pseudoroegneria spicata-Danthonia californica-Achnatherum lemmonii) grasslands; and tufted hairgrass (Deschampsia cespitosa)-California oatgrass vernally flooded prairie.

Cities have become more committed to ecological restoration and management activities in urban natural areas.

This guidebook describes major biological and physical attributes of the 433-ha (1,070-ac) Grayback Glades Research Natural Area.

This guidebook describes major biological and physical attributes of the 3531-ha (8,725-ac) Fish Creek Rim Research Natural Area located within the Northern Basin and Range ecoregion and managed by the Bureau of Land Management, Lakeview District (USDI BLM 2003).

This guidebook describes major biological and physical attributes of the 155-ha (384-ac) Lost Lake Research Natural Area (RNA), in Jackson County, Oregon. The RNA has been designated because it contains examples of a landslide-dammed lake; and a low-elevation lake with aquatic beds and fringing marsh, surrounded by mixed-conifer forest (ONHAC 2010).

This book is intended to help people better understand how wood energy is helping to revitalize rural Alaskan communities by reducing energy costs, creating jobs, and helping to educate the next generation.

Due to the COVID-19 CORONAVIRUS pandemic, JACK BARTON ENTERTAINMENT (JBE) has decided to postpone the first JBE TRIPLE A SUMMITFEST in BOULDER, CO, scheduled for AUGUST 4-7, until AUGUST … more

Summer is going to look a bit different in the Corridor this year as many, long-cherished events are being canceled or postponed. And the organizations that run those events want you to know they...

Summer is going to look a bit different in the Corridor this year as many, long-cherished events are being canceled or postponed. And the organizations that run those events want you to know they aren’t any more happy about it than you are.

The organizers of these events are having to make unprecedented, tough decisions.

“Cancellation is not a good word in our business,” said Chuck Swanson, Building a Legacy executive director of Hancher. “It is something that we really don’t want to do and it takes a lot for us to come to that.

“We live for the live performance and bringing the artists and audiences together. That’s the happiest time for me, so none of these decisions have been easy.”

Hancher has had to cancel numerous upcoming events in the past few months that would have brought to Iowa City in artists from all over the country and the world. It also is holding off announcing its upcoming season — which it typically would be doing at this time of year.

this isn’t something the staff has faced since the floods of 2008 and because they book events so far in advance they are confronting additional challenges.

“You know there’s so much that goes into a show before it happens,” Swanson said. “I just think of all the anticipation, booking the artists, advancing the show, setting ticket prices, advertising and then ticket sales.

“It’s like a farmer who does all this work to get his crops ready and then at the end of the season ends up with nothing to harvest.”

He noted Hancher has been reaching out to its booked performers and, in some cases, have had performers reach out to them to cancel upcoming shows.

The significant time and resources that go into planning large-scale events is the main factor in necessitating cancellation discussions and decisions at many organizations.

“Many logistical items have to be coordinated, from renting shuttles to scheduling volunteers and staff. Initial planning for some events begins as early as 12 to 18 months in advance and proceeds all the way up to the day of the event,” said John Myers, Indian Creek Nature Center executive director.

Citing the center’s annual Maple Syrup Festival, he noted food represents a significant cost and often cannot be saved or reused.

“We have had to be mindful of the financial resources available to us and ensure that we wisely manage those to ensure (the center) can emerge from this pandemic as a functioning and healthy organization,” he said.

“None of the decisions to cancel events or how to handle subsequent financial losses are easy and they challenge everyone,” Myers added. “As our whole lives have been upended, it makes even the simplest of decisions harder and that takes an impact on morale.”

He acknowledged staff members aren’t the only ones feeling the strain.

“We have a significant core of volunteers who are no longer able to give their time, which also creates a strain on morale and increases the amount of work that needs to be done when we return,” he pointed out.

Another primary factor is what is allowed and considered safe by the city, state and Iowa Department of Public Health.

“At this point, only allowing groups of 10 or less is a far cry from the thousands or people we usually see at the Iowa Arts Festival,” said Lisa Barnes, executive director of Summer of the Arts in Iowa City, which produces the Iowa Arts Festival.

“The governor has announced that reopening the state will be done in stages, and based on what we’ve found from other events around the country, concerts and large festivals will be the last to open,” he noted.

Summer of the Arts announced just last week that the Iowa Arts Festival would not take place this year, a month in advance of the event.

“We needed to make a decision so that we can move forward with alternative plans,” Barnes said, noting the organization has had questions about the Iowa City Jazz Festival, scheduled for July 3 through 5 and added a decision regarding that festival and July programming will be made by mid-late May.

“We also needed to make the decision far enough out to be able to work with our performers and cancel the agreements,” she said.

On Wednesday, Gov. Kim Reynolds loosened some but not all of the social-distancing restrictions for the remaining 22 counties she had put in place.

Discussions about the future of these events have been happening for weeks for many organizations, highlighting they are not taken lightly.

Carissa Johnson, executive director of the Cedar Rapids Freedom Festival, said conversations about the future of this year’s event started in mid-March, right around the time the Cedar Rapids SaPaDaPaSo Parade announced its cancellation for 2020.

“We plan year ’round for the two- to three-week festival,” Johnson explained.

“Our planning really ramps up in April and May, and we have many more costs associated with producing the festival the closer we get to the start. In order to protect our time and resources, we elected to cancel before we had more costs and variables to consider.”

As for who is making the final decision, organizations said many stakeholders are involved. Barnes said the decision on the Iowa City Arts Festival, for example, included staff, the board of directors, festival planning committees, the city of Iowa City and Johnson County Public Health, along with input from some of the vendors, artists and performers.

Tapping into experts in those public health field has been key as well.

“We have these assets, people, at the University (of Iowa), that have been really helpful as we make these decisions about canceling and as we prepare to think about reopening,” Hancher’s Swanson said.

The Freedom Festival include staff and board members in discussions, with recommendations from Linn County Public Health and the city of Cedar Rapids, factoring in the health, safety and well-being of the community.

“We are just as heartbroken as the rest of the community, but this decision was to protect our community as much as possible,” Swanson said.

“This community is a family and we will all get through this together and come back stronger next year.”

Myers noted organizations such as the Indian Creek Nature Center are also rely on advice from national associations, such as the American Alliance of Museums, and discussions among the leadership of many local cultural groups.

“For many events, we have also reached out to participants to gather their input and comfort level of attending once we are able to reopen,” Myers said.

The financial effects of having to cancel is stressful for organizations, too.

“Financially, this has been a hard time for the Nature Center to endure,” Myers pointed out. “We’ve had over 100 different programs, events and facility rentals canceled between March 15 and April 30, and our losses are currently over $250,000. As we approach the summer, there are a number of other events we continue to review, including our popular summer camps.”

The Nature Center has postponed a national conference to be held there in September — due to indications of low participation — for peers from around the nation who run not-for-profit and government nature centers.

“We are losing thousands of dollars in vendor fees and sales receipts because we had to cancel,” said Barnes, of Summer of the Arts.

“We have sponsors tied to certain events, like the Iowa Arts Festival, that in some cases want to carry over their support to next year, which impacts our fundraising for this year and next.”

She noted her group already has been made aware of funding that won’t be coming in from some sponsors next year due to the financial impact those organizations are facing as well.

And that can be tough.

“When we cancel, our whole staff is involved — from the box audience and public engagement folks to the technical production team and our front-of-house staff,” Swanson said.

“Our communication is key in talking through it all and then sharing clear messages with our audiences, especially in terms of refunds. But we’ve been encouraged by so many generous friends of Hancher donating their ticket purchase price back to us.”

While disappointment still is thick in the air, organizations don’t plan to abandon their missions and is keeping an eye on serving the public.

“This is a challenging time for everyone, and our board and staff is committed to finding creative and non-traditional solutions to ensure the Freedom Festival’s return,” Johnson said. “The community and our stakeholders have been tremendous supports of the Freedom Festival and we believe they will continue to do so in the future.

“We ask for understanding and patience as we try to navigate this crisis and what we can still provide for our community.”

Freedom Festival buttons will be sold this year as they’ve already been made, and “It’s a way the community can show their support,” Johnson said.

Barnes agreed and noted the Iowa Arts Festival committee is working on ways to support the performers, artists and vendors they had scheduled by trying to develop some virtual opportunities for engagement.

While the show, or events, might not go on, organizers said they very much want to remain connected to their audiences and attendees.

“I want to make sure everybody knows we care about them and that we’re trying to find ways to stay connected because I think we’re all in this together and the arts are one of the best ways for people to get through difficult times,” Swanson said.

Myers agreed.

“Indian Creek Nature Center will be ready to welcome guests and visitors back to our events as soon as we are able to do so safely,” he said.

“In the meantime, we hope everyone finds peace in nature by taking a hike or bike ride, having a picnic or just enjoying time outside.”

The pandemic has changed the daily lives of everyone. How we work, how we shop, and how we interact with each other are all shifting. Comparing life as it is now with how it used to be can lead to sadness or despair and what's called "ambiguous loss."

Tags: Jukebox, music, neural network, OpenAI

It’s A Living Mural for X-Games China & Innersect

The pandemic has changed the daily lives of everyone. How we work, how we shop, and how we interact with each other are all shifting. Comparing life as it is now with how it used to be can lead to sadness or despair and what's called "ambiguous loss."

In this work, we investigate multi-task learning as a way of pre-training models for classification tasks in digital pathology. It is motivated by the fact that many small and medium-size datasets have been released by the community over the years whereas there is no large scale dataset similar to ImageNet in the domain. We first assemble and transform many digital pathology datasets into a pool of 22 classification tasks and almost 900k images. Then, we propose a simple architecture and training scheme for creating a transferable model and a robust evaluation and selection protocol in order to evaluate our method. Depending on the target task, we show that our models used as feature extractors either improve significantly over ImageNet pre-trained models or provide comparable performance. Fine-tuning improves performance over feature extraction and is able to recover the lack of specificity of ImageNet features, as both pre-training sources yield comparable performance.

A standard approach to evaluating language models analyzes how models assign probabilities to valid versus invalid syntactic constructions (i.e. is a grammatical sentence more probable than an ungrammatical sentence). Our work uses ambiguous relative clause attachment to extend such evaluations to cases of multiple simultaneous valid interpretations, where stark grammaticality differences are absent. We compare model performance in English and Spanish to show that non-linguistic biases in RNN LMs advantageously overlap with syntactic structure in English but not Spanish. Thus, English models may appear to acquire human-like syntactic preferences, while models trained on Spanish fail to acquire comparable human-like preferences. We conclude by relating these results to broader concerns about the relationship between comprehension (i.e. typical language model use cases) and production (which generates the training data for language models), suggesting that necessary linguistic biases are not present in the training signal at all.

Deep neural networks have exhibited promising performance in image super-resolution (SR). Most SR models follow a hierarchical architecture that contains both the cell-level design of computational blocks and the network-level design of the positions of upsampling blocks. However, designing SR models heavily relies on human expertise and is very labor-intensive. More critically, these SR models often contain a huge number of parameters and may not meet the requirements of computation resources in real-world applications. To address the above issues, we propose a Hierarchical Neural Architecture Search (HNAS) method to automatically design promising architectures with different requirements of computation cost. To this end, we design a hierarchical SR search space and propose a hierarchical controller for architecture search. Such a hierarchical controller is able to simultaneously find promising cell-level blocks and network-level positions of upsampling layers. Moreover, to design compact architectures with promising performance, we build a joint reward by considering both the performance and computation cost to guide the search process. Extensive experiments on five benchmark datasets demonstrate the superiority of our method over existing methods.

In-band full-duplex systems can transmit and receive information simultaneously on the same frequency band. However, due to the strong self-interference caused by the transmitter to its own receiver, the use of non-linear digital self-interference cancellation is essential. In this work, we describe a hardware architecture for a neural network-based non-linear self-interference (SI) canceller and we compare it with our own hardware implementation of a conventional polynomial based SI canceller. In particular, we present implementation results for a shallow and a deep neural network SI canceller as well as for a polynomial SI canceller. Our results show that the deep neural network canceller achieves a hardware efficiency of up to $312.8$ Msamples/s/mm$^2$ and an energy efficiency of up to $0.9$ nJ/sample, which is $2.1 imes$ and $2 imes$ better than the polynomial SI canceller, respectively. These results show that NN-based methods applied to communications are not only useful from a performance perspective, but can also be a very effective means to reduce the implementation complexity.

The standard training algorithm in neural machine translation (NMT) suffers from exposure bias, and alternative algorithms have been proposed to mitigate this. However, the practical impact of exposure bias is under debate. In this paper, we link exposure bias to another well-known problem in NMT, namely the tendency to generate hallucinations under domain shift. In experiments on three datasets with multiple test domains, we show that exposure bias is partially to blame for hallucinations, and that training with Minimum Risk Training, which avoids exposure bias, can mitigate this. Our analysis explains why exposure bias is more problematic under domain shift, and also links exposure bias to the beam search problem, i.e. performance deterioration with increasing beam size. Our results provide a new justification for methods that reduce exposure bias: even if they do not increase performance on in-domain test sets, they can increase model robustness to domain shift.

Deep learning-based models, such as convolutional neural networks, have advanced various segments of computer vision. However, this technology is rarely applied to seismic shot gather noise localization problem. This letter presents an investigation on the effectiveness of a multi-scale feature-fusion-based network for seismic shot-gather noise localization. Herein, we describe the following: (1) the construction of a real-world dataset of seismic noise localization based on 6,500 seismograms; (2) a multi-scale feature-fusion-based detector that uses the MobileNet combined with the Feature Pyramid Net as the backbone; and (3) the Single Shot multi-box detector for box classification/regression. Additionally, we propose the use of the Focal Loss function that improves the detector's prediction accuracy. The proposed detector achieves an AP@0.5 of 78.67\% in our empirical evaluation.

We present a new system, EEV, for verifying binarized neural networks (BNNs). We formulate BNN verification as a Boolean satisfiability problem (SAT) with reified cardinality constraints of the form $y = (x_1 + cdots + x_n le b)$, where $x_i$ and $y$ are Boolean variables possibly with negation and $b$ is an integer constant. We also identify two properties, specifically balanced weight sparsity and lower cardinality bounds, that reduce the verification complexity of BNNs. EEV contains both a SAT solver enhanced to handle reified cardinality constraints natively and novel training strategies designed to reduce verification complexity by delivering networks with improved sparsity properties and cardinality bounds. We demonstrate the effectiveness of EEV by presenting the first exact verification results for $ell_{infty}$-bounded adversarial robustness of nontrivial convolutional BNNs on the MNIST and CIFAR10 datasets. Our results also show that, depending on the dataset and network architecture, our techniques verify BNNs between a factor of ten to ten thousand times faster than the best previous exact verification techniques for either binarized or real-valued networks.

In recent years there has been a burgeoning interest in the use of computational methods to distinguish between elicited speech samples produced by patients with dementia, and those from healthy controls. The difference between perplexity estimates from two neural language models (LMs) - one trained on transcripts of speech produced by healthy participants and the other trained on transcripts from patients with dementia - as a single feature for diagnostic classification of unseen transcripts has been shown to produce state-of-the-art performance. However, little is known about why this approach is effective, and on account of the lack of case/control matching in the most widely-used evaluation set of transcripts (DementiaBank), it is unclear if these approaches are truly diagnostic, or are sensitive to other variables. In this paper, we interrogate neural LMs trained on participants with and without dementia using synthetic narratives previously developed to simulate progressive semantic dementia by manipulating lexical frequency. We find that perplexity of neural LMs is strongly and differentially associated with lexical frequency, and that a mixture model resulting from interpolating control and dementia LMs improves upon the current state-of-the-art for models trained on transcript text exclusively.

In recent years, deep neural networks have been applied to obtain high performance of prediction, classification, and pattern recognition. However, the weights in these deep neural networks are difficult to be explained. Although a linear regression method can provide explainable results, the method is not suitable in the case of input interaction. Therefore, an explainable deep neural network (ExpDNN) with explainable layers is proposed to obtain explainable results in the case of input interaction. Three cases were given to evaluate the proposed ExpDNN, and the results showed that the absolute value of weight in an explainable layer can be used to explain the weight of corresponding input for feature extraction.