Rename Nerf Sample to Mobile NeRF (KhronosGroup#865)

* Rename files and readme

* Fix wrong mobile nerf header include

* Fix sample building issue with wrong invocation name

* Update readme with a better description of traditional NeRF vs the mobile one

* Update nerf shader path to mobile_nerf

antora/modules/ROOT/nav.adoc

@@ -105,7 +105,7 @@
 * xref:samples/tooling/README.adoc[Tooling samples]
 ** xref:samples/tooling/profiles/README.adoc[Profiles]
 * xref:samples/general/README.adoc[General samples]
-** xref:samples/general/nerf/README.adoc[NeRF]
+** xref:samples/general/mobile_nerf/README.adoc[Mobile NeRF]
 * xref:docs/README.adoc[General documentation]
 ** xref:docs/build.adoc[Build guide]
 ** xref:docs/memory_limits.adoc[Memory limits]

samples/CMakeLists.txt

@@ -131,7 +131,7 @@ set(ORDER_LIST
     "hpp_swapchain_images"
 
     #General Samples
-    "nerf")
+    "mobile_nerf")
 
 # Orders the sample ids by the order list above
 set(ORDERED_LIST)

samples/README.adoc

@@ -512,6 +512,6 @@ The library defines a common baseline of features, extensions, etc.
 
 The goal of these samples is to demonstrate different techniques or showcase complex scenarios that doesn't necessarily fit any of the main categories.
 
-=== xref:./general/nerf/README.adoc[NeRF]
+=== xref:./general/mobile_nerf/README.adoc[Mobile NeRF]
 
 A Neural Radiance Field synthesizer sample, based on textured polygons.

samples/general/nerf/CMakeLists.txt → samples/general/mobile_nerf/CMakeLists.txt

@@ -23,6 +23,6 @@ add_sample_with_tags(
     ID ${FOLDER_NAME}
     CATEGORY ${CATEGORY_NAME}
     AUTHOR "Qualcomm"
-    NAME "NeRF"
-    DESCRIPTION "A Neural Radiance Field synthesizer sample, based on textured polygons"
+    NAME "Mobile NeRF"
+    DESCRIPTION "A Mobile Neural Radiance Field synthesizer sample, based on textured polygons"
 )

samples/general/nerf/README.adoc → samples/general/mobile_nerf/README.adoc

@@ -17,14 +17,24 @@
 -
 ////
 
-= NeRF
+= Mobile NeRF
 
 ifdef::site-gen-antora[]
-TIP: The source for this sample can be found in the https://github.com/KhronosGroup/Vulkan-Samples/tree/main/samples/general/nerf[Khronos Vulkan samples github repository].
+TIP: The source for this sample can be found in the https://github.com/KhronosGroup/Vulkan-Samples/tree/main/samples/general/mobile_nerf[Khronos Vulkan samples github repository].
 endif::[]
 
 This sample is a modified version of the https://mobile-nerf.github.io/[Mobile Nerf] developed by Google. 
 It's based on its original https://github.com/google-research/jax3d/tree/main/jax3d/projects/mobilenerf[source code] but optimized for Vulkan.
+This is a different version from traditional NeRF rendering, which normally requires tracing rays (usually done via ray-marching) and querying a MLP multiple times for each ray. These many queries result in non-interactive frame rates on most of the GPUs.
+The mobile version uses the rasterization pipeline to render the final image; this is done via a triangle mesh and a feature texture, where each of its visible pixels are run through a small MLP (executed in the fragment shader) that converts the feature data and view direction to the corresponding output pixel color. This technique enables interactive FPS even on mobile GPUs (thus the name).
+
+just rendering the standard NeRF in realtime is not feasible on commodity hardware. Rendering
+views from a trained NeRF requires querying a multi-layer
+perception (MLP) hundreds of times per ray. It requires
+about 100 teraflops to render a single 800∗800 frame, which
+results in a best-case rendering time of 10 seconds per frame
+on a NVIDIA RTX 2080 GPU with full GPU utilization.
+
 
 == Description: [https://mobile-nerf.github.io/[Mobile Nerf]]
 Neural Radiance Fields (NeRFs) have demonstrated amazing ability to synthesize images of 3D scenes from novel views. 

samples/general/nerf/nerf.cpp → samples/general/mobile_nerf/mobile_nerf.cpp

@@ -15,7 +15,7 @@
  * limitations under the License.
  */
 
-#include "nerf.h"
+#include "mobile_nerf.h"
 #include "glm/gtx/matrix_decompose.hpp"
 #include "gltf_loader.h"
 #include "platform/filesystem.h"
@@ -92,17 +92,17 @@ void camera_set_look_at(vkb::Camera &camera, const glm::vec3 look, const glm::ve
 
 }        // namespace
 
-Nerf::Nerf()
+MobileNerf::MobileNerf()
 {
-	title = "NeRF";
+	title = "Mobile NeRF";
 	// SPIRV 1.4 requires Vulkan 1.1
 	set_api_version(VK_API_VERSION_1_1);
 	add_device_extension(VK_KHR_SPIRV_1_4_EXTENSION_NAME);
 	// Required by VK_KHR_spirv_1_4
 	add_device_extension(VK_KHR_SHADER_FLOAT_CONTROLS_EXTENSION_NAME);
 }
 
-Nerf::~Nerf()
+MobileNerf::~MobileNerf()
 {
 	if (device)
 	{
@@ -175,7 +175,7 @@ Nerf::~Nerf()
 	}
 }
 
-void Nerf::read_json_map()
+void MobileNerf::read_json_map()
 {
 	std::string assetBase = vkb::fs::path::get(vkb::fs::path::Type::Assets);
 	LOGI("Base assets path: {}", assetBase);
@@ -373,34 +373,34 @@ void Nerf::read_json_map()
 	}
 }
 
-void Nerf::load_shaders()
+void MobileNerf::load_shaders()
 {
 	// Loading first pass shaders
 	if (use_deferred)
 	{
 		// Loading first pass shaders
-		shader_stages_first_pass[0] = load_shader("nerf/raster.vert", VK_SHADER_STAGE_VERTEX_BIT);
+		shader_stages_first_pass[0] = load_shader("mobile_nerf/raster.vert", VK_SHADER_STAGE_VERTEX_BIT);
 		shader_stages_first_pass[1] = load_shader(
-		    using_original_nerf_models[0] ? "nerf/raster.frag" : "nerf/raster_morpheus.frag",
+		    using_original_nerf_models[0] ? "mobile_nerf/raster.frag" : "mobile_nerf/raster_morpheus.frag",
 		    VK_SHADER_STAGE_FRAGMENT_BIT);
 
 		// Loading second pass shaders
-		shader_stages_second_pass[0] = load_shader("nerf/quad.vert", VK_SHADER_STAGE_VERTEX_BIT);
+		shader_stages_second_pass[0] = load_shader("mobile_nerf/quad.vert", VK_SHADER_STAGE_VERTEX_BIT);
 		shader_stages_second_pass[1] = load_shader(
-		    using_original_nerf_models[0] ? "nerf/mlp.frag" : "nerf/mlp_morpheus.frag",
+		    using_original_nerf_models[0] ? "mobile_nerf/mlp.frag" : "mobile_nerf/mlp_morpheus.frag",
 		    VK_SHADER_STAGE_FRAGMENT_BIT);
 	}
 	else
 	{
 		// Loading one pass shaders
-		shader_stages_first_pass[0] = load_shader("nerf/raster.vert", VK_SHADER_STAGE_VERTEX_BIT);
+		shader_stages_first_pass[0] = load_shader("mobile_nerf/raster.vert", VK_SHADER_STAGE_VERTEX_BIT);
 		shader_stages_first_pass[1] = load_shader(
-		    using_original_nerf_models[0] ? "nerf/merged.frag" : "nerf/merged_morpheus.frag",
+		    using_original_nerf_models[0] ? "mobile_nerf/merged.frag" : "mobile_nerf/merged_morpheus.frag",
 		    VK_SHADER_STAGE_FRAGMENT_BIT);
 	}
 }
 
-bool Nerf::prepare(const vkb::ApplicationOptions &options)
+bool MobileNerf::prepare(const vkb::ApplicationOptions &options)
 {
 	read_json_map();
 
@@ -488,18 +488,18 @@ bool Nerf::prepare(const vkb::ApplicationOptions &options)
 	return true;
 }
 
-bool Nerf::resize(const uint32_t width, const uint32_t height)
+bool MobileNerf::resize(const uint32_t width, const uint32_t height)
 {
 	ApiVulkanSample::resize(width, height);
 	rebuild_command_buffers();
 	return true;
 }
 
-void Nerf::request_gpu_features(vkb::PhysicalDevice &gpu)
+void MobileNerf::request_gpu_features(vkb::PhysicalDevice &gpu)
 {
 }
 
-void Nerf::render(float delta_time)
+void MobileNerf::render(float delta_time)
 {
 	if (!prepared)
 	{
@@ -514,7 +514,7 @@ inline uint32_t aligned_size(uint32_t value, uint32_t alignment)
 	return (value + alignment - 1) & ~(alignment - 1);
 }
 
-void Nerf::setup_attachment(VkFormat format, VkImageUsageFlags usage, FrameBufferAttachment &attachment)
+void MobileNerf::setup_attachment(VkFormat format, VkImageUsageFlags usage, FrameBufferAttachment &attachment)
 {
 	if (attachment.image != VK_NULL_HANDLE)
 	{
@@ -600,7 +600,7 @@ void Nerf::setup_attachment(VkFormat format, VkImageUsageFlags usage, FrameBuffe
 	VK_CHECK(vkCreateSampler(get_device().get_handle(), &samplerCreateInfo, 0, &attachment.sampler));
 }
 
-void Nerf::setup_nerf_framebuffer_baseline()
+void MobileNerf::setup_nerf_framebuffer_baseline()
 {
 	if (use_deferred)
 	{
@@ -671,7 +671,7 @@ void Nerf::setup_nerf_framebuffer_baseline()
 	}
 }
 
-void Nerf::update_descriptor_sets_baseline()
+void MobileNerf::update_descriptor_sets_baseline()
 {
 	for (int i = 0; i < nerf_framebuffers.size(); i++)
 	{
@@ -702,7 +702,7 @@ void Nerf::update_descriptor_sets_baseline()
 	}
 }
 
-void Nerf::build_command_buffers()
+void MobileNerf::build_command_buffers()
 {
 	if (use_native_screen_size)
 	{
@@ -712,7 +712,7 @@ void Nerf::build_command_buffers()
 	build_command_buffers_baseline();
 }
 
-void Nerf::build_command_buffers_baseline()
+void MobileNerf::build_command_buffers_baseline()
 {
 	// In case the screen is resized, need to update the storage image size and descriptor set
 	// Note that the texture_rendered image has already been recreated at this point
@@ -830,7 +830,7 @@ void Nerf::build_command_buffers_baseline()
 	}
 }
 
-void Nerf::load_scene(int model_index, int sub_model_index, int models_entry)
+void MobileNerf::load_scene(int model_index, int sub_model_index, int models_entry)
 {
 	Model &model = models[models_entry];
 
@@ -895,7 +895,7 @@ void Nerf::load_scene(int model_index, int sub_model_index, int models_entry)
 	}
 }
 
-void Nerf::create_descriptor_pool()
+void MobileNerf::create_descriptor_pool()
 {
 	if (use_deferred)
 	{
@@ -923,7 +923,7 @@ void Nerf::create_descriptor_pool()
 	}
 }
 
-void Nerf::create_pipeline_layout_fist_pass()
+void MobileNerf::create_pipeline_layout_fist_pass()
 {
 	// First Pass Descriptor set and layout
 
@@ -950,7 +950,7 @@ void Nerf::create_pipeline_layout_fist_pass()
 	VK_CHECK(vkCreatePipelineLayout(get_device().get_handle(), &pipeline_layout_create_info, nullptr, &pipeline_first_pass_layout));
 }
 
-void Nerf::create_pipeline_layout_baseline()
+void MobileNerf::create_pipeline_layout_baseline()
 {
 	// Second Pass Descriptor set and layout
 
@@ -974,7 +974,7 @@ void Nerf::create_pipeline_layout_baseline()
 	VK_CHECK(vkCreatePipelineLayout(get_device().get_handle(), &pipeline_layout_create_info, nullptr, &pipeline_layout_baseline));
 }
 
-void Nerf::create_descriptor_sets_first_pass(Model &model)
+void MobileNerf::create_descriptor_sets_first_pass(Model &model)
 {
 	int numDescriptorPerModel = use_deferred ? 1 : nerf_framebuffers.size();
 	model.descriptor_set_first_pass.resize(numDescriptorPerModel);
@@ -1024,7 +1024,7 @@ void Nerf::create_descriptor_sets_first_pass(Model &model)
 	}
 }
 
-void Nerf::create_descriptor_sets_baseline()
+void MobileNerf::create_descriptor_sets_baseline()
 {
 	descriptor_set_baseline.resize(nerf_framebuffers.size());
 
@@ -1066,7 +1066,7 @@ void Nerf::create_descriptor_sets_baseline()
 	}
 }
 
-void Nerf::prepare_pipelines()
+void MobileNerf::prepare_pipelines()
 {
 	VkPipelineInputAssemblyStateCreateInfo input_assembly_state = vkb::initializers::pipeline_input_assembly_state_create_info(VK_PRIMITIVE_TOPOLOGY_TRIANGLE_LIST, 0, VK_FALSE);
 
@@ -1160,7 +1160,7 @@ void Nerf::prepare_pipelines()
 	}
 }
 
-void Nerf::create_static_object_buffers(int model_index, int sub_model_index, int models_entry)
+void MobileNerf::create_static_object_buffers(int model_index, int sub_model_index, int models_entry)
 {
 	LOGI("Creating static object buffers");
 	Model &model              = models[models_entry];
@@ -1203,7 +1203,7 @@ void Nerf::create_static_object_buffers(int model_index, int sub_model_index, in
 	LOGI("Done Creating static object buffers");
 }
 
-void Nerf::create_uniforms()
+void MobileNerf::create_uniforms()
 {
 	uniform_buffers.resize(model_path.size());
 	weights_buffers.resize(model_path.size());
@@ -1233,7 +1233,7 @@ void Nerf::create_uniforms()
 	update_uniform_buffers();
 }
 
-void Nerf::initialize_mlp_uniform_buffers(int model_index)
+void MobileNerf::initialize_mlp_uniform_buffers(int model_index)
 {
 	std::string assetBase   = vkb::fs::path::get(vkb::fs::path::Type::Assets);
 	std::string mlpJsonPath = assetBase + model_path[model_index] + "mlp.json";
@@ -1388,7 +1388,7 @@ void Nerf::initialize_mlp_uniform_buffers(int model_index)
 	}
 }
 
-void Nerf::update_uniform_buffers()
+void MobileNerf::update_uniform_buffers()
 {
 	assert(uniform_buffers[0]);
 
@@ -1415,7 +1415,7 @@ void Nerf::update_uniform_buffers()
 	}
 }
 
-void Nerf::prepare_instance_data()
+void MobileNerf::prepare_instance_data()
 {
 	auto &ii = instancing_info;
 
@@ -1472,7 +1472,7 @@ void Nerf::prepare_instance_data()
 	device->get_fence_pool().wait();
 }
 
-void Nerf::draw()
+void MobileNerf::draw()
 {
 	ApiVulkanSample::prepare_frame();
 
@@ -1486,7 +1486,7 @@ void Nerf::draw()
 	ApiVulkanSample::submit_frame();
 }
 
-void Nerf::create_texture(int model_index, int sub_model_index, int models_entry)
+void MobileNerf::create_texture(int model_index, int sub_model_index, int models_entry)
 {
 	// Set up the input texture image
 
@@ -1504,7 +1504,7 @@ void Nerf::create_texture(int model_index, int sub_model_index, int models_entry
 	LOGI("Done Creating feature texture 0");
 }
 
-void Nerf::create_texture_helper(std::string texturePath, Texture_Input &texture_input)
+void MobileNerf::create_texture_helper(std::string texturePath, Texture_Input &texture_input)
 {
 	// Copy data to an optimal tiled image
 	// This loads the texture data into a host local buffer that is copied to the optimal tiled image on the device
@@ -1623,7 +1623,7 @@ void Nerf::create_texture_helper(std::string texturePath, Texture_Input &texture
 	VK_CHECK(vkCreateSampler(get_device().get_handle(), &samplerCreateInfo, 0, &texture_input.sampler));
 }
 
-void Nerf::update_render_pass_nerf_forward()
+void MobileNerf::update_render_pass_nerf_forward()
 {
 	// For merged shaders, we need 2 attachments (as opposed to 5)
 	// 0: Depth attachment
@@ -1677,7 +1677,7 @@ void Nerf::update_render_pass_nerf_forward()
 	VK_CHECK(vkCreateRenderPass(device->get_handle(), &render_pass_create_info, nullptr, &render_pass_nerf));
 }
 
-void Nerf::update_render_pass_nerf_baseline()
+void MobileNerf::update_render_pass_nerf_baseline()
 {
 	std::array<VkAttachmentDescription, 5> attachments = {};
 	// Color attachment 1
@@ -1815,7 +1815,7 @@ void Nerf::update_render_pass_nerf_baseline()
 	VK_CHECK(vkCreateRenderPass(device->get_handle(), &render_pass_create_info, nullptr, &render_pass_nerf));
 }
 
-std::unique_ptr<vkb::VulkanSample> create_nerf()
+std::unique_ptr<vkb::VulkanSample> create_mobile_nerf()
 {
-	return std::make_unique<Nerf>();
+	return std::make_unique<MobileNerf>();
 }

samples/general/nerf/nerf.h → samples/general/mobile_nerf/mobile_nerf.h

@@ -35,11 +35,11 @@ class Camera;
 }        // namespace sg
 }        // namespace vkb
 
-class Nerf : public ApiVulkanSample
+class MobileNerf : public ApiVulkanSample
 {
   public:
-	Nerf();
-	~Nerf() override;
+	MobileNerf();
+	~MobileNerf() override;
 	void request_gpu_features(vkb::PhysicalDevice &gpu) override;
 	void render(float delta_time) override;
 	bool prepare(const vkb::ApplicationOptions &options) override;
@@ -240,4 +240,4 @@ class Nerf : public ApiVulkanSample
 	bool     use_native_screen_size = false;
 };
 
-std::unique_ptr<vkb::VulkanSample> create_nerf();
+std::unique_ptr<vkb::VulkanSample> create_mobile_nerf();